def getBlastScoreRatios(FASTAfile, allelescore, queryDef, databasePath, queryProteomeName, referenceGenomeArray, referenceCDS, isXML):
alleleProt=''
proteome=""
countP=0
countCDS=0
if isXML == 'True':
blast_out_file = 'BLASTresults.xml'
cline = NcbiblastpCommandline(query=queryProteomeName, db=databasePath, out=blast_out_file, outfmt=5, num_alignments=7000, num_descriptions=7000)
print 'BSR:'
#print cline
blast_records = runBlastParser(cline,blast_out_file, False)
startTime = datetime.now()
ToNewAllele = parseBLASTRecordsXML(blast_records, allelescore, queryDef, referenceGenomeArray, referenceCDS)
print 'CheckResults:' + str(datetime.now() - startTime)
else:
blast_out_file = 'BLASTresults.tab'
cline = NcbiblastpCommandline(query=queryProteomeName, db=databasePath, out=blast_out_file, outfmt=6, num_alignments=7000, num_descriptions=7000)
print 'BSR:'
blast_records = runBlastParserTAB(cline,blast_out_file, False)
startTime = datetime.now()
ToNewAllele = parseBLASTRecordsTAB(blast_records, allelescore, queryDef, referenceGenomeArray, referenceCDS)
print 'CheckResults:' + str(datetime.now() - startTime)
os.remove(queryProteomeName)
return ToNewAllele
def reDogetBlastScoreRatios(genefile,basepath,alleleI,allelescores2,newGene_Blast_DB_name,alleleList2,picklepath):
gene_fp = HTSeq.FastaReader(genefile)
alleleProt=''
alleleI+=1
proteinfastaPath=genefile
print ("Re-starting Blast alleles at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_out_file2 = os.path.join(basepath,'blastdbs/temp.xml')
cline = NcbiblastpCommandline(query=proteinfastaPath, db=newGene_Blast_DB_name, evalue=0.001, out=blast_out_file2, outfmt=5)
allelescore=0
blast_records = runBlastParser(cline,blast_out_file2, proteinfastaPath)
print ("Blasted alleles at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
found =False
for blast_record in blast_records:
for alignment in blast_record.alignments:
for match in alignment.hsps:
allelescores2.append(int(match.score))
var=[alleleI,allelescores2]
with open(picklepath,'wb') as f:
currentCDSDict = pickle.dump(var, f)
return int(alleleI),allelescores2,alleleList2
def getOwnBlastScore(FASTAfile, databasePath, queryProteomeName, numberOfLocus, blastResultsPath, LocusToUse, queryFile):
allelescores = []
alleleNumbers = {}
sameAlleles = {}
prevAlleleName = {}
databasePath, isEmpty, proteinsToQueryFile, queryAlleleList, prevAlleleName = CreateQueryDatabase(FASTAfile, databasePath,queryProteomeName)
if isEmpty:
return allelescores, isEmpty, proteinsToQueryFile, alleleNumbers, sameAlleles, prevAlleleName
blast_out_file = blastResultsPath + '/' + numberOfLocus + '_BLASTresults.xml'
cline = NcbiblastpCommandline(query=queryProteomeName, db=databasePath, out=blast_out_file, outfmt=5, num_alignments=7000, num_descriptions=7000)
#print cline
allelescore=0
blast_records = runBlastParser(cline,blast_out_file, False)
allelescores, alleleList, alleleNumbers, sameAlleles = parseOwnBLASTRecordsAndDuplicates(blast_records, FASTAfile, queryAlleleList)
proteinsToQueryFile = translateAlleleList(alleleList, queryProteomeName, LocusToUse, queryFile)
os.remove(databasePath+ ".pin")
os.remove(databasePath+ ".phr")
os.remove(databasePath+ ".psq")
os.remove(databasePath+ "_blast.log")
os.remove(blast_out_file)
return allelescores, isEmpty, proteinsToQueryFile, alleleNumbers, sameAlleles, prevAlleleName
def getBlastScoreRatios(allelescore, alleleList, databasePath,
queryProteomeName, referenceGenomeArray, referenceCDS,
bestmatches, referenceCDSsequences, referenceFileName,
countNumberOfGenomes, blastResultsPath, LocusToUse):
alleleProt = ''
proteome = ""
countP = 0
countCDS = 0
blast_out_file = blastResultsPath + countNumberOfGenomes + '_BLASTresults.xml'
cline = NcbiblastpCommandline(query=queryProteomeName,
db=databasePath,
out=blast_out_file,
outfmt=5,
num_alignments=7000,
num_descriptions=7000)
#print cline
blast_records = runBlastParser(cline, blast_out_file, False)
resultsList, addNewAlleles = parseBLASTRecordsXML(
blast_records, allelescore, alleleList, referenceGenomeArray,
referenceCDS, bestmatches, referenceCDSsequences, referenceFileName,
LocusToUse)
os.remove(blast_out_file)
return resultsList, addNewAlleles
def getOwnBlastScore(FASTAfile):
gene_fp = HTSeq.FastaReader(FASTAfile)
#alleleI=0
names=""
alleleProt=''
proteome=""
for allele in gene_fp: #new db for each allele to blast it against himself
try:
x = str(translateSeq(allele.seq))
except:
continue
#print str(allele.name)
#names=allele.name.split("|")[3]
#print allele.seq
alleleProt+=">"+str(allele.name)+"\n"+x+"\n"
proteome+=">"+str(allele.name)+"\n"+x+"\n"
with open(pathRef+'allAllelesAA.fasta', "wb") as f:
f.write(alleleProt)
with open(pathRef+nameOrg+'proteome.fasta', "wb") as v:
v.write(proteome)
Gene_Blast_DB_name = Create_Blastdb(pathRef+'allAllelesAA.fasta',1,True)
# --- get BLAST score ratio --- #
cline = NcbiblastpCommandline(query=pathRef+nameOrg+'proteome.fasta', db=name, out=blast_out_file, outfmt=5, num_alignments=7000, num_descriptions=7000)
#print cline
allelescore=0
blast_records = runBlastParser(cline,blast_out_file, alleleProt)
allelescores={}
for blast_record in blast_records:
found=False
for alignment in blast_record.alignments:
if found is False:
#print blast_record.query, alignment.hit_def
for match in alignment.hsps:
#print alignment.hit_def
#print "---------------------"
#print alignment.hit_def
#print blast_record.query
#print alignment.hit_def
try:
if allelescores[str(alignment.hit_def)] < match.score:
allelescores[str(alignment.hit_def)] = int(match.score)
break
except KeyError:
allelescores[str(alignment.hit_def)] = int(match.score)
break
else:
break
#print allelescores
#for i in allelescores:
#hitsName.append(str(i)+";"+str(allelescores[i])+";")
#hitsName.sort(key=Align_sort_key)
#print hitsName
#return alleleI,allelescores,Gene_Blast_DB_name
#print alleleI
#print len(allelescores)
return allelescores
def main():
parser = argparse.ArgumentParser(description="Given an ffn file, recovers the genes that are not paralogs and have a size bigger than the g parameter provided")
parser.add_argument('-i', nargs='?', type=str, help='ffn file', required=True)
parser.add_argument('-g', nargs='?', type=int, help='int minimum size', required=True)
args = parser.parse_args()
genes = args.i
sizethresh = args.g
gene_fp = HTSeq.FastaReader(genes)
geneFile = os.path.abspath( genes )
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 )
geneF = os.path.splitext( geneFile )[0]
blast_out_file = geneF + '.xml'
# list of results - the output of the function
resultsList = []
# ------------------------------ RUNNING BLAST ------------------------------ #
cline = NcbiblastnCommandline(query=geneFile, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, geneFile)
paralogs=[]
for blast_record in blast_records:
try:
alignment=blast_record.alignments[1]
paralogs.append( alignment.hit_def)
except:
continue
pathfiles=os.path.dirname(geneFile)
pathfiles=pathfiles+"/"
print pathfiles
g_fp = HTSeq.FastaReader( genes )
removedparalogs=0
removedsize=0
for contig in g_fp:
name = contig.name+" "+contig.descr
if name not in paralogs:
if int(len(contig.seq))>sizethresh:
namefile=contig.name
namefile=namefile.replace("|","_")
with open(pathfiles+namefile+".fasta", "wb") as f:
f.write(">1\n"+contig.seq+"\n")
else:
removedsize+=1
else:
print name
removedparalogs+=1
print "Removed %s paralog genes" % str(removedparalogs)
print "Removed %s because of size :" % str(removedsize)
def reDogetBlastScoreRatios(genefile,basepath,alleleI,allelescores2,newGene_Blast_DB_name,alleleList2): gene_fp = HTSeq.FastaReader(genefile) #alleleI=0 #allelescores=[] alleleProt='' #alleleList=[] """for allele in gene_fp: #new db for each allele to blast it against himself print allele alleleI+=1 genome=-1 alleleList2.append(allele.seq) translatedSequence,x,y=translateSeq(allele.seq) print translatedSequence alleleProt+=">"+str(alleleI)+"\n"+str(translatedSequence+"\n")""" alleleI+=1 proteinfastaPath=genefile print proteinfastaPath blast_out_file2 = os.path.join(basepath,'blastdbs/temp.xml') #with open(proteinfastaPath, "wb") as f: # f.write(alleleProt) #Gene_Blast_DB_name = Create_Blastdb( proteinfastaPath, 1, True ) # --- get BLAST score ratio --- # cline = NcbiblastpCommandline(query=proteinfastaPath, db=newGene_Blast_DB_name, evalue=0.001, out=blast_out_file2, outfmt=5) #print cline allelescore=0 blast_records = runBlastParser(cline,blast_out_file2, proteinfastaPath) found =False for blast_record in blast_records: found=False print blast_record #print blast_record.header #print blast_record.alignments[0] for alignment in blast_record.alignments: print alignment,alignment.hsps if found is False: #print blast_record.query, alignment.hit_def for match in alignment.hsps: print match #print "---------------------" if(int(alignment.hit_def)== int(blast_record.query)): #print match allelescores2.append(int(match.score)) found=True break else: break #print allelescores2, alleleList2 return alleleI,allelescores2,alleleList2
def getBlastScoreRatios(genefile,basepath):
gene_fp = HTSeq.FastaReader(genefile)
alleleI=0
allelescores=[]
alleleProt=''
alleleList=[]
for allele in gene_fp: #new db for each allele to blast it against himself
alleleI+=1
genome=-1
alleleList.append(allele.seq)
translatedSequence,x,y=translateSeq(allele.seq)
alleleProt+=">"+str(alleleI)+"\n"+str(translatedSequence+"\n")
#basepath="./blastdbs/temp"+str(os.path.basename(genefile))
#if not os.path.exists(basepath):
# os.makedirs(basepath)
proteinfastaPath=os.path.join(basepath,str(os.path.basename(genefile)+'_protein.fasta'))
blast_out_file = os.path.join(basepath,"blastdbs/"+os.path.basename(genefile) + '.xml')
with open(proteinfastaPath, "wb") as f:
f.write(alleleProt)
print ("Starting Blast alleles at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
Gene_Blast_DB_name = Create_Blastdb( proteinfastaPath, 1, True )
print proteinfastaPath
# --- get BLAST score ratio --- #
cline = NcbiblastpCommandline(query=proteinfastaPath, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
#print cline
allelescore=0
print ("Parse bsr blast at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_records = runBlastParser(cline,blast_out_file, alleleProt)
print ("Blasted alleles at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
for blast_record in blast_records:
found=False
for alignment in blast_record.alignments:
if found is False:
#print blast_record.query, alignment.hit_def
for match in alignment.hsps:
#print "---------------------"
if(int(alignment.hit_def)== int(blast_record.query)):
#print match
allelescores.append(int(match.score))
found=True
break
else:
break
#print allelescores
return alleleI,allelescores,Gene_Blast_DB_name,alleleList
def reDogetBlastScoreRatios(genefile, basepath, alleleI, allelescores2,
newGene_Blast_DB_name, alleleList2, picklepath,
verbose, blastPath, listAllelesNames):
if verbose:
def verboseprint(*args):
for arg in args:
print(arg),
print
else:
verboseprint = lambda *a: None # do-nothing function
#gene_fp = HTSeq.FastaReader(genefile)
alleleProt = ''
proteinfastaPath = genefile
verboseprint("Starting Blast of new alleles to calculate BSR at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_out_file2 = os.path.join(basepath, 'blastdbs/temp.xml')
cline = NcbiblastpCommandline(cmd=blastPath,
query=proteinfastaPath,
db=newGene_Blast_DB_name,
evalue=0.001,
out=blast_out_file2,
outfmt=5,
num_threads=1)
allelescore = 0
blast_records = runBlastParser(cline, blast_out_file2)
verboseprint("Blasted alleles at : " + time.strftime("%H:%M:%S-%d/%m/%Y"))
found = False
matchscore = 0
for blast_record in blast_records:
for alignment in blast_record.alignments:
for match in alignment.hsps:
matchscore = int(match.score)
allelescores2[alleleI] = matchscore
with open(picklepath, 'wb') as f:
pickle.dump(allelescores2, f)
return allelescores2, alleleList2, listAllelesNames
def BLASTp(queryFile, dbName, blast_out_path, queryNames, sequenceLengths):
blast_out_file = os.path.join(blast_out_path,'blastOut.xml')
cline = NcbiblastpCommandline(query=queryFile, db=dbName, out=blast_out_file, outfmt=5)
blast_records = runBlastParser(cline,blast_out_file, "")
matchGene = ''
score = -1
for blast_record in blast_records:
queryGeneIndex = queryNames.index(blast_record.query.strip('|'))
querySequenceLength = sequenceLengths[queryGeneIndex]
for alignment in blast_record.alignments:
for match in alignment.hsps:
identity_length_ratio = float(match.identities)/float(querySequenceLength)
if identity_length_ratio >= 0.8:
if score < match.score:
matchGene = alignment.hit_def
score = match.score
return matchGene
def getBlastScoreRatios(allelescore, alleleList, databasePath, queryProteomeName, referenceGenomeArray, referenceCDS, bestmatches, referenceCDSsequences,referenceFileName, countNumberOfGenomes, blastResultsPath, LocusToUse):
alleleProt=''
proteome=""
countP=0
countCDS=0
blast_out_file = blastResultsPath + countNumberOfGenomes + '_BLASTresults.xml'
cline = NcbiblastpCommandline(query=queryProteomeName, db=databasePath, out=blast_out_file, outfmt=5, num_alignments=7000, num_descriptions=7000)
#print cline
blast_records = runBlastParser(cline,blast_out_file, False)
resultsList, addNewAlleles = parseBLASTRecordsXML(blast_records, allelescore, alleleList, referenceGenomeArray, referenceCDS, bestmatches, referenceCDSsequences, referenceFileName, LocusToUse)
os.remove(blast_out_file)
return resultsList, addNewAlleles
def getOwnBlastScore(FASTAfile, databasePath, queryProteomeName, numberOfLocus,
blastResultsPath, LocusToUse, queryFile):
allelescores = []
alleleNumbers = {}
sameAlleles = {}
prevAlleleName = {}
databasePath, isEmpty, proteinsToQueryFile, queryAlleleList, prevAlleleName = CreateQueryDatabase(
FASTAfile, databasePath, queryProteomeName)
if isEmpty:
return allelescores, isEmpty, proteinsToQueryFile, alleleNumbers, sameAlleles, prevAlleleName
blast_out_file = blastResultsPath + '/' + numberOfLocus + '_BLASTresults.xml'
cline = NcbiblastpCommandline(query=queryProteomeName,
db=databasePath,
out=blast_out_file,
outfmt=5,
num_alignments=7000,
num_descriptions=7000)
#print cline
allelescore = 0
blast_records = runBlastParser(cline, blast_out_file, False)
allelescores, alleleList, alleleNumbers, sameAlleles = parseOwnBLASTRecordsAndDuplicates(
blast_records, FASTAfile, queryAlleleList)
proteinsToQueryFile = translateAlleleList(alleleList, queryProteomeName,
LocusToUse, queryFile)
os.remove(databasePath + ".pin")
os.remove(databasePath + ".phr")
os.remove(databasePath + ".psq")
os.remove(databasePath + "_blast.log")
os.remove(blast_out_file)
return allelescores, isEmpty, proteinsToQueryFile, alleleNumbers, sameAlleles, prevAlleleName
def getBlastScoreRatios(genefile): gene_fp = HTSeq.FastaReader(genefile) alleleI=0 allelescores=[] alleleProt='' for allele in gene_fp: #new db for each allele to blast it against himself alleleI+=1 genome=-1 alleleProt+=">"+str(alleleI)+"\n"+str(translateSeq(allele.seq)+"\n") basepath="./blastdbs/temp"+str(os.path.basename(genefile)) if not os.path.exists(basepath): os.makedirs(basepath) with open(basepath+'/protein.fasta', "wb") as f: f.write(alleleProt) Gene_Blast_DB_name = Create_Blastdb( basepath+'/protein.fasta', 1, True ) # --- get BLAST score ratio --- # cline = NcbiblastpCommandline(query=basepath+'/protein.fasta', db=Gene_Blast_DB_name, evalue=0.001, out=basepath+'protein.xml', outfmt=5) #print cline allelescore=0 blast_records = runBlastParser(cline,basepath+'protein.xml', alleleProt) for blast_record in blast_records: found=False for alignment in blast_record.alignments: if found is False: #print blast_record.query, alignment.hit_def for match in alignment.hsps: #print "---------------------" if(int(alignment.hit_def)== int(blast_record.query)): #print match allelescores.append(int(match.score)) found=True break else: break #print allelescores return alleleI,allelescores,Gene_Blast_DB_name
def reDogetBlastScoreRatios(genefile, basepath, alleleI, allelescores2,
newGene_Blast_DB_name, alleleList2, picklepath):
gene_fp = HTSeq.FastaReader(genefile)
alleleProt = ''
alleleI += 1
proteinfastaPath = genefile
print("Re-starting Blast alleles at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_out_file2 = os.path.join(basepath, 'blastdbs/temp.xml')
cline = NcbiblastpCommandline(query=proteinfastaPath,
db=newGene_Blast_DB_name,
evalue=0.001,
out=blast_out_file2,
outfmt=5)
allelescore = 0
blast_records = runBlastParser(cline, blast_out_file2, proteinfastaPath)
print("Blasted alleles at : " + time.strftime("%H:%M:%S-%d/%m/%Y"))
found = False
for blast_record in blast_records:
for alignment in blast_record.alignments:
for match in alignment.hsps:
allelescores2.append(int(match.score))
var = [alleleI, allelescores2]
with open(picklepath, 'wb') as f:
currentCDSDict = pickle.dump(var, f)
return int(alleleI), allelescores2, alleleList2
def callAlleles(argumentList):
geneFile = argumentList[0]
genomesList = argumentList[1]
listOfProts = argumentList[2]
listAllCDS = argumentList[3]
#print geneFile
gene_fp = HTSeq.FastaReader(geneFile)
alleleI = 0
#inverted=False
#orderedAlleleNames=[]
resultsList = []
i = 0
perfectMatchIdAllele=[]
bestmatch=[0,0,False,'',0] #score, score ratio, perfectmatch, key name of the DNA sequence string, allele ID
allelescores=[]
alleleI,allelescores,Gene_Blast_DB_name=getBlastScoreRatios(geneFile)
genome=-1
for protList in listOfProts:
#alleleI = 0
#alleleProt=''
#for allele in gene_fp: #new db for each allele to blast it against himself
# alleleI+=1
# alleleProt+=">"+str(alleleI)+"\n"+str(translateSeq(allele.seq)+"\n")
basepath="./blastdbs/temp"+str(os.path.basename(geneFile))
#if not os.path.exists(basepath):
# os.makedirs(basepath)
#with open(basepath+'/protein.fasta', "wb") as f:
# f.write(alleleProt)
#Gene_Blast_DB_name = Create_Blastdb( basepath+'/protein.fasta', 1, True )
genome+=1
with open(basepath+'/proteinList.fasta', "wb") as f:
f.write(protList)
#Gene_Blast_DB_name = Create_Blastdb( './temp/proteinList.fasta', 1, True )
cline = NcbiblastpCommandline(query=basepath+'/proteinList.fasta', db=Gene_Blast_DB_name, evalue=0.001, out=basepath+'proteinList.xml', outfmt=5)
#print cline
blast_records = runBlastParser(cline, basepath+'proteinList.xml', basepath+'/proteinList.fasta')
for blast_record in blast_records:
for alignment in blast_record.alignments:
#print alignment
#print alignment.hsps
#print alignment.hit_id
#print alignment.hit_def
#print alignment.title
for match in alignment.hsps:
#print blast_record.query
#print match
#print alleleI, len(allelescores)
scoreRatio=float(match.score)/float(allelescores[int(alignment.hit_def)-1])
#print scoreRatio
#print alignment.hit_def
cdsStrName=blast_record.query
if(scoreRatio == 1 and bestmatch[2] is False):
bestmatch=[match.score,scoreRatio,True,cdsStrName,int(alignment.hit_def)]
#print alignment
#print match
elif(scoreRatio == 1 and match.score>bestmatch[0]):
bestmatch=[match.score,scoreRatio,True,cdsStrName,int(alignment.hit_def)]
#print match
elif(match.score>bestmatch[0] and scoreRatio>0.4 and scoreRatio>bestmatch[1] and bestmatch[2] is False):
#print match.query
#print match.sbjct
#print allelescores
bestmatch=[match.score,scoreRatio,False,cdsStrName,int(alignment.hit_def)]
#print match
#print bestmatch
if bestmatch[0]==0:
#if no best match was found
###################
# LOCUS NOT FOUND #
###################
resultsList.append('LNF3:-1') # append result to the list of results
perfectMatchIdAllele.append('LNF')
#printinfo(genomeFile,geneFile)
print "Locus not found, no matches \n"
elif bestmatch[2] is True:
#if a perfect match was found
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
perfectMatchIdAllele.append(str(bestmatch[4]))
resultsList.append('EXC:' + str(bestmatch[4]) )
else:
#######################
# ADD INFERRED ALLELE # # a new allele
#######################
#print "infered allele has location : "+(CDSType)
#printinfo(genomeFile,geneFile)
tagAux='INF'
perfectMatchIdAllele.append( tagAux +"-"+str(alleleI+1))
print "New allele! Adding allele "+ tagAux + str(alleleI+1) +" to the database\n"
resultsList.append( tagAux + str(alleleI+1) )
#orderedAlleleNames.append('allele_' + str(alleleI) + '_' + tagAux[:-1] +"_" +str(os.path.basename(genomeFile)))
# --- add the new allele to the gene fasta --- #
fG = open( geneFile, 'a' )
fG.write('>allele_' + str(alleleI+1) + '_' + tagAux[:-1] +"_" + str(os.path.basename(genomesList[genome])) + '\n')
#print alleleStr
listOfCDS=listAllCDS[genome]
#print listOfCDS
fG.write( listOfCDS[">"+bestmatch[3]] + '\n')
fG.close()
#alleleI += 1
# --- remake blast DB --- #
Gene_Blast_DB_name = Create_Blastdb( basepath+'/protein.fasta', 1, True )
alleleI,allelescores,Gene_Blast_DB_name=getBlastScoreRatios(geneFile)
#x=y
shutil.rmtree(basepath)
final = (resultsList,perfectMatchIdAllele)
#return (resultsList)
return final
def callAlleles(argumentList):
geneFile = argumentList[0]
genomesList = argumentList[1]
listOfCDSDicts = argumentList[2]
listOfGenomesDict = argumentList[3]
gene_fp = HTSeq.FastaReader(geneFile)
geneDict = {}
alleleI = 1
inverted=False
orderedAlleleNames=[]
biggestAllelelen=0
for allele in gene_fp:
if allele.seq in geneDict:
print "\nWARNING: this file contains a repeated allele, it should be checked. Ignoring it now!\n", geneFile
else:
if len(allele.seq)>biggestAllelelen:
biggestAllelelen=len(allele.seq)
orderedAlleleNames.append(allele.name)
geneDict[ allele.seq ] = alleleI
alleleI += 1
#print geneDict
#print orderedAlleleNames
# --- make 1st blast DB --- #
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 )
geneF = os.path.splitext( geneFile )[0]
blast_out_file = geneF + '.xml'
# list of results - the output of the function
resultsList = []
i = 0
perfectMatchIdAllele=[]
for genomeFile in genomesList:
#print geneDict
currentCDSDict = listOfCDSDicts[i]
currentGenomeDict = listOfGenomesDict[i]
#print genomeFile
#print resultsList
#print geneDict
#print orderedAlleleNames
i+=1 # it has to be incremented here
if genomeFile[-1] == '\n':
genomeFile = genomeFile[:-1]
# ------------------------------ RUNNING BLAST ------------------------------ #
cline = NcbiblastnCommandline(query=genomeFile, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, genomeFile)
# ------ DETERMINING BEST MATCH ------ #
# bestMatch = ['rec.query','hsp', lenRatio]
bestMatch = ['','', 0]
bestMatchContig=''
bestMatchContigLen=''
bestalignlen=0
perfectMatch=False
bmAlleleLen2=0
bmAllele=''
#noAlignment=False
for blast_record in blast_records:
# --- the LNF cases are now called outside de loop --- #
#print blast_record
if perfectMatch==True:
break
try:
#print blast_record.alignments
hspC = blast_record.alignments[0]
if bestMatch[0] == '' and bestMatch[1] == '':
bestMatch[0] = blast_record.query
bestMatch[1] = hspC
except IndexError:
continue
# --- the contig tag is used in the progigal function --- #
contigTag = blast_record.query
# --- brute force parsing of the contig tag - better solution is advisable --- #
j=0
for l in contigTag:
if l == ' ':
break
j+=1
contigTag = contigTag[:j]
contigLen = blast_record.query_letters
#print blast_record.query_id
# --- iterating over all the results to determine the best match --- #
for alignment in blast_record.alignments:
index=orderedAlleleNames.index(alignment.hit_def)
#print alignment.hit_def
for k, v in geneDict.iteritems():
if v == index+1:
bmAlleleLen2= len(k)
if perfectMatch:
break
for match in alignment.hsps:
#print match
scoreRatio = float(match.score) / float(bmAlleleLen2)
#print alignment.hit_def
#print match.identities
#print bmAlleleLen2
#print
#print match.identities
#print len(match.sbjct)
if (int(match.identities)==int(bmAlleleLen2) and int(match.identities)==int(len(match.sbjct)) and "N" not in match.query ):
index=orderedAlleleNames.index(alignment.hit_def)
bmAlleleLen= len(geneDict.keys()[index])
lenratio=float(len(match.query))/float(bmAlleleLen)
bestMatch = [blast_record.query, match, scoreRatio, alignment.hit_def,lenratio,bmAlleleLen]
bestMatchContig=contigTag
perfectMatch=True
index=orderedAlleleNames.index(alignment.hit_def)
bmAlleleLen= len(geneDict.keys()[index])
break
elif scoreRatio > bestMatch[2]:
index=orderedAlleleNames.index(alignment.hit_def)
bmAllele=geneDict.keys()[index]
bmAlleleLen= len(bmAllele)
lenratio=float(len(match.query))/float(bmAlleleLen)
bestMatch = [blast_record.query, match, scoreRatio, alignment.hit_def,lenratio,bmAlleleLen]
bestMatchContig=contigTag
bestMatchContigLen=blast_record.query_letters
if match.sbjct_start > match.sbjct_end:
inverted=True
#print match.query
bestalignlen=alignment.length
if perfectMatch==True:
break
# ---------- ALLELE CALLING AFTER DETERMINING BEST MATCH ---------- #
try:
#print bestMatch[0]
match = bestMatch[1]
#print match.query
geneLen = bestMatch[5]
alleleStr = match.query
nIdentities = match.identities
idPercent = float(nIdentities) / float(geneLen)
scoreRatio = bestMatch[2]
lenRatio = bestMatch[4]
#print perfectMatch
#print "\nContig best/exact match is :"
#print bestMatchContig +"\n"
except:
resultsList.append('LNF3:-1') # append result to the list of results
perfectMatchIdAllele.append('LNF')
printinfo(genomeFile,geneFile)
print "Locus not found \n"
continue
#TODO check identities >0.8
if perfectMatch is True:
#TODO perfect match to top
if match.sbjct_start > match.sbjct_end:
alleleStr = reverseComplement(alleleStr)
#TODO test replace -
#alleleStr = alleleStr.replace('-', '')
alleleNumber = geneDict[ alleleStr ]
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
if "_" in bestMatch[3]:
a=bestMatch[3].split("_")
perfectMatchIdAllele.append(a[1])
else:
perfectMatchIdAllele.append(bestMatch[3])
resultsList.append('EXC:' + str(alleleNumber) )
###################
# LOCUS NOT FOUND #
###################
#elif bestMatch[0] == '':
# resultsList.append('LNF:-1') # append result to the list of results
# perfectMatchIdAllele.append('LNF')
# printinfo(genomeFile,geneFile)
# print "Locus not found \n"
elif bestMatch[0] != '' and perfectMatch is not True:
###########################
# LOCUS ON THE CONTIG TIP #
###########################
#if match.query_start == 1 or bestMatchContigLen <= match.query_end:
## TODO-
## 1 - LOT5 match.query_start ==1 and match.length < match.subj.length (allele length) alignement length
## 2 - LOT 3' match.query_end == match.query.length (contig length) and match.length < contig length (allele length??)
## 3 - LOT SC bestMatchContigLen <= allele length
if match.query_start ==1 and len(match.query) < geneLen:
resultsList.append('LOT5:-1')
perfectMatchIdAllele.append('LOT5')
printinfo(genomeFile,geneFile)
print "Locus is on the 5' tip of the contig \n"
elif match.query_end == bestMatchContigLen and len(match.query) < bestMatchContigLen:
resultsList.append('LOT3:-1')
perfectMatchIdAllele.append('LOT3')
printinfo(genomeFile,geneFile)
print "Locus is on the 3' tip of the contig \n"
elif bestMatchContigLen <= geneLen:
resultsList.append('LOTSC:-1')
perfectMatchIdAllele.append('LOTSC')
printinfo(genomeFile,geneFile)
#print match.query_start
print "Locus is bigger than the contig \n"
elif 'N' in alleleStr:
#TODO gravar para ficheiro
#####################
# ALLELE NOT FOUND # # N base found!
#####################
geneFile2= os.path.splitext(geneFile)[0] + "LNFN.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+"\n")
f.write((alleleStr) +"\n")
resultsList.append('LNFN:-1')
perfectMatchIdAllele.append('LNFN')
printinfo(genomeFile,geneFile)
print "LNFN, contains N bases! \n"
else:
# ------------------------------------------------------------------------------------------------------- #
# #
# USING PRODIGAL TO TRY TO EXTEND CDS #
# #
# ------------------------------------------------------------------------------------------------------- #
CDSType=''
extended, strCDS, CDSType = extendCDS(bestMatchContig, currentCDSDict, match.query_start, match.query_end, currentGenomeDict, geneLen)
# --- if it was possible to extend it using prodigal --- #
#print extended
#print strCDS
#print CDSType
if extended :
alleleStr = strCDS
lenRatio = float(len(strCDS)) / float(geneLen)
#print alleleStr
#print lenRatio
elif not extended and biggestAllelelen > geneLen:
extended, strCDS, CDSType = extendCDS(bestMatchContig, currentCDSDict, match.query_start, match.query_end, currentGenomeDict, biggestAllelelen)
if extended :
alleleStr = strCDS
lenRatio = float(len(strCDS)) / float(geneLen)
else:
alleleStr = alleleStr.replace('-', '')
else:
# --- removing gaps '-' --- #
#print alleleStr
alleleStr = alleleStr.replace('-', '')
# --- continuing the allele calling --- #
#print geneDict
#print alleleStr
# --- it might be needed to obtain the reverse complement of the allele string --- #
if match.sbjct_start > match.sbjct_end:
alleleStr = reverseComplement(alleleStr)
if alleleStr in geneDict:
alleleNumber = geneDict[ alleleStr ]
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
perfectMatchIdAllele.append(alleleNumber)
resultsList.append('EXC:' + str(alleleNumber) )
else:
isUndefined = False
#print geneDict.keys()[0]
defAllele=''
defAlleleName=''
for k in geneDict.keys():
if alleleStr in k:
defAllele=k
#print alleleStr
isUndefined = True
defAlleleName=geneDict.get(k)
break
if extended and isUndefined and idPercent > 0.8 and ((int(len(match.query))==int(len(defAllele)) or int(len(match.query))==int(len(defAllele))+1 or int(len(match.query))==int(len(defAllele))-1)) :
#extended allele to compare may be different from the allele to compare from bm
alleleStr=match.query
alleleStr = alleleStr.replace('-', '')
if match.sbjct_start > match.sbjct_end: #### - error??
alleleStr = reverseComplement(alleleStr)
if int(len(alleleStr))==int(len(defAllele)): # se o match for do mesmo tamanho que o alello
tagAux = 'NA1:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA1-"+str(alleleI))
elif int(len(alleleStr))==int(len(defAllele))-1 : # se o match tiver uma base a mais que o alelo
tagAux = 'NA2:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA2-"+str(alleleI))
else: #se o match tiver uma base a menos que o alelo
tagAux = 'NA3:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA3-"+str(alleleI))
print "New allele found! Adding allele "+ tagAux + str(alleleI) +" to the database"
geneDict[alleleStr] = alleleI
resultsList.append( tagAux + str(alleleI) )
orderedAlleleNames.append('allele_' + str(alleleI) + '_' + tagAux[:-1] + str(os.path.basename(genomeFile)))
# --- add the new allele to the gene fasta --- #
fG = open( geneFile, 'a' )
fG.write('>allele_' + str(alleleI) + '_' + tagAux[:-1] + str(os.path.basename(genomeFile)) + '\n')
fG.write( alleleStr + '\n')
fG.close()
alleleI += 1
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 )
elif not extended and idPercent > 0.8 and ((int(len(match.query))==int(geneLen) or int(len(match.query))==int(geneLen)+1 or int(len(match.query))==int(geneLen)-1)) :
alleleStr=match.query
alleleStr = alleleStr.replace('-', '')
if match.sbjct_start > match.sbjct_end: #### - error??
alleleStr = reverseComplement(alleleStr)
if int(len(alleleStr))==int(geneLen): # se o match for do mesmo tamanho que o alello
tagAux = 'NA4:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA4-"+str(alleleI))
elif int(len(alleleStr))==int(geneLen)-1 : # se o match tiver uma base a mais que o alelo
tagAux = 'NA5:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA5-"+str(alleleI))
else: #se o match tiver uma base a menos que o alelo
tagAux = 'NA6:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA6-"+str(alleleI))
print "New allele found! Adding allele "+ tagAux + str(alleleI) +" to the database"
geneDict[alleleStr] = alleleI
resultsList.append( tagAux + str(alleleI) )
orderedAlleleNames.append('allele_' + str(alleleI) + '_' + tagAux[:-1] + str(os.path.basename(genomeFile)))
# --- add the new allele to the gene fasta --- #
fG = open( geneFile, 'a' )
fG.write('>allele_' + str(alleleI) + '_' + tagAux[:-1] + str(os.path.basename(genomeFile)) + '\n')
fG.write( alleleStr + '\n')
fG.close()
alleleI += 1
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 )
elif isUndefined:
####################
# UNDEFINED ALLELE # # it is contained in another allele
####################
alleleStr=match.query
#if match.sbjct_start > match.sbjct_end: #### - error
#alleleStr = reverseComplement(alleleStr)
resultsList.append('UND:-1')
perfectMatchIdAllele.append("undefined allele")
printinfo(genomeFile,geneFile)
print "Undefined allele \n"
geneFile2= os.path.splitext(geneFile)[0] + "undefined.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
#f.write(">BlastBestMatch"+str(defAlleleName)+"\n")
#f.write((alleleStr)+"\n")
f.write(">Allele"+str(defAlleleName)+"\n")
f.write((defAllele)+"\n")
else:
if not extended :
if lenRatio < 0.5:
###############
# SMALL MATCH #
###############
resultsList.append('SAC:-1') # don't know what 'SAC' stands for
perfectMatchIdAllele.append('small match')
printinfo(genomeFile,geneFile)
print "lower than 50% match \n"
elif lenRatio < 0.8 and idPercent < 0.5:
#####################
# INCOMPLETE ALLELE # # it was not possible to extend it to at least 80% of the length of the gene
#####################
resultsList.append('INC:-1')
perfectMatchIdAllele.append('allele incomplete')
printinfo(genomeFile,geneFile)
print "Incomplete allele\n"
else:
##################
# LNF WTFFF #
##################
geneFile2= os.path.splitext(geneFile)[0] + "LNF2.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
f.write(">Allele\n")
f.write((bmAllele)+"\n")
resultsList.append('LNF2')
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("LNF2")
print "Not extended and no allele found"
else:
#######################
# ADD INFERRED ALLELE # # a new allele that was extended with prodigal
#######################
if(CDSType=='larger than match'):
tagAux = 'INF1:'
elif(CDSType=='start codon inside match'):
tagAux = 'INF2:'
elif(CDSType=='early stop codon in match'):
tagAux = 'INF3:'
elif(CDSType=='same size as allele'):
tagAux = 'INF4:'
else:
tagAux = 'INF5:'
print "infered allele has location : "+(CDSType)
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append( tagAux +"-"+str(alleleI))
print "New allele Infered with prodigal! Adding allele "+ tagAux + str(alleleI) +" to the database\n"
geneDict[alleleStr] = alleleI
resultsList.append( tagAux + str(alleleI) )
orderedAlleleNames.append('allele_' + str(alleleI) + '_' + tagAux[:-1] + str(os.path.basename(genomeFile)))
# --- add the new allele to the gene fasta --- #
fG = open( geneFile, 'a' )
fG.write('>allele_' + str(alleleI) + '_' + tagAux[:-1] + str(os.path.basename(genomeFile)) + '\n')
#print alleleStr
fG.write( alleleStr + '\n')
fG.close()
alleleI += 1
# --- remake blast DB --- #
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 )
final = (resultsList,perfectMatchIdAllele)
#return (resultsList)
return final
def main():
parser = argparse.ArgumentParser(description="Given an ffn file, recovers the genes that are not paralogs and have a size bigger than the g parameter provided")
parser.add_argument('-i', nargs='?', type=str, help='ffn file', required=True)
parser.add_argument('-g', nargs='?', type=int, help='int minimum size', required=True)
args = parser.parse_args()
genes = args.i
sizethresh = args.g
passSteps = False
#translate to protein and create new file
abspath=os.path.abspath(genes)
filename=os.path.basename(genes)
abspath=abspath.replace(filename,'')
proteinfile=os.path.join(abspath,'proteins.fasta')
geneDict = {}
protDict={}
orderedprotDict=collections.OrderedDict()
alreadyIn=[]
totalgenes=0
repeatedgenes=0
smallgenes=0
if not passSteps:
print "not passing steps"
with open(proteinfile, "wb") as f:
g_fp = HTSeq.FastaReader( genes )
totalgenes+=1
for gene in g_fp:
dnaseq= str(gene.seq)
protseq,x,y=translateSeq(dnaseq)
if len(protseq)>1:
if str(protseq) in alreadyIn:
repeatedgenes+=1
elif len(str(protseq))<67:
smallgenes+=1
else:
alreadyIn.append(str(protseq))
protname=">"+str(gene.name)+"\n"
f.write(protname+str(protseq)+"\n")
protDict[protname] = str(protseq)
geneDict[str(gene.name)] = gene.seq
else:
print gene.name
orderedprotList=[]
orderedprotList=sorted(protDict.items(), key=lambda x: len(x[1]), reverse=True)
i=0
while i < len(orderedprotList):
elem=orderedprotList[i]
orderedprotDict[elem[0]] = elem[1]
i+=1
#print orderedprotDict
print str(repeatedgenes) + " repeated genes out of "+ str(totalgenes)
print str(smallgenes) + " small genes out of "+ str(totalgenes)
print "protein file created"
# first step - remove genes contained in other genes or 100% equal genes
# list of results - the output of the function
resultsList = []
auxDict={}
g_fp = HTSeq.FastaReader( proteinfile )
g=0
j=0
print "Checking if proteins are equal or substring of others..."
# for each gene from all the annotated genes - starting with an empty dictionary, only add a new gene if the "to be added gene" is not contained or equal to a gene already added to the dictionary
auxprot=[]
for elem in orderedprotDict.items():
contained=False
prot=str(elem[1])
if any(prot in x for x in auxprot):
g+=1
contained=True
else:
auxDict[elem[1]] = elem[0]
auxprot.append(str(elem[1]))
print str(j)+ " out of " + str(len(orderedprotDict) )
j+=1
#print "____" +str(j)
print "%s genes are contained in other genes" % (g)
#overwrite the original file, obtaining a new file with unique genes
with open(proteinfile, "wb") as f:
allsequences=''
for k,v in auxDict.iteritems():
allsequences+=v+k+"\n"
f.write(allsequences)
else:
totalgenes=0
smallgenes=0
g_fp = HTSeq.FastaReader( genes )
totalgenes+=1
for gene in g_fp:
dnaseq= str(gene.seq)
protseq,x,y=translateSeq(dnaseq)
if len(protseq)>1:
if str(protseq) in alreadyIn:
repeatedgenes+=1
#print gene.name + " already saved "
elif len(str(protseq))<67:
smallgenes+=1
else:
alreadyIn.append(str(protseq))
protname=">"+str(gene.name)+"\n"
#print protseq
protDict[protname] = str(protseq)
geneDict[str(gene.name)] = gene.seq
else:
print gene.name
geneFile = os.path.abspath( proteinfile )
print proteinfile
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1, True )
geneF = os.path.splitext( geneFile )[0]
blast_out_file = geneF + '.xml'
# ------------------------------ RUNNING BLAST ------------------------------ #
cline = NcbiblastpCommandline(query=geneFile, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, geneFile)
toRemove=[]
genesToKeep=[]
log=["removed\tcause\texplanation"]
for blast_record in blast_records:
allelename=blast_record.query
allelename=allelename.split(" ")
allelename=allelename[0]
alleleLength=len(geneDict[allelename])
try:
#if gene A is not on the toRemove list yet, add to genesToKeep list
if str(blast_record.query) not in toRemove:
genesToKeep.append(blast_record.query)
i=0
#if first alignement is not against self, gene B is bigger than gene A and very simillar - remove gene A from genesToKeep and add gene B instead
if not str(blast_record.query) == str((blast_record.alignments[0]).hit_def):
genesToKeep.remove(str(blast_record.query))
toRemove.append(str(blast_record.query))
log.append(str(blast_record.query)+"\t"+str((blast_record.alignments[0]).hit_def)+"\t"+"2 is first best match")
#if gene B is not on the toRemove list, add to genesToKeep list
if str((blast_record.alignments[0]).hit_def) not in toRemove:
genesToKeep.append(str((blast_record.alignments[0]).hit_def))
raise
selfblastscore=(((blast_record.alignments[0]).hsps)[0]).score
while i<len(blast_record.alignments):
align=blast_record.alignments[i]
match=(align.hsps)[0]
scoreRatio=float(match.score)/float(selfblastscore)
alleleLength2=len(geneDict[str(align.hit_def)])
#if good match and gene B not in toremove list
if(scoreRatio>0.6 and not str(align.hit_def) == str(blast_record.query) and str(align.hit_def) not in toRemove):
#if gene B is bigger than gene A, keep bigger gene B
if alleleLength2>alleleLength :
genesToKeep.append(str(align.hit_def))
genesToKeep.remove(str(blast_record.query))
toRemove.append(str(blast_record.query))
log.append(str(blast_record.query)+"\t"+str(align.hit_def)+"\t"+"2 is bigger and bsr >0.6")
raise
#else add gene B to toremove list
elif str(align.hit_def) in genesToKeep:
genesToKeep.remove(str(align.hit_def))
toRemove.append(str(align.hit_def))
log.append(str(align.hit_def)+"\t"+str(blast_record.query)+"\t"+"2 is bigger and bsr >0.6")
i+=1
#else gene A is on toRemove list, add all similar genes (not in genesToKeep) list to the toRemove list
else:
i=0
selfblastscore=0
for align in blast_record.alignments:
if not (str(align.hit_def) == str(blast_record.query)):
selfblastscore=((align.hsps)[0]).score
print "gene "+str(align.hit_def)+" is bigger than gene "+str(blast_record.query)
raise
while i<len(blast_record.alignments):
align=blast_record.alignments[i]
match=(align.hsps)[0]
scoreRatio=float(match.score)/float(selfblastscore)
if align.hit_def not in genesToKeep and not str(align.hit_def) == str(blast_record.query) and scoreRatio>0.6 :
toRemove.append(align.hit_def)
log.append(str(align.hit_def)+"\t"+str(blast_record.query)+"\t"+"2 was on the removed list and bsr >0.6")
else:
pass
i+=1
except Exception as e:
#print e
pass
with open("logfile.txt", "wb") as f:
for elem in log:
f.write(str(elem)+"\n")
genesToKeep=list(set(genesToKeep))
toRemove=list(set(toRemove))
s = set(toRemove)
notcommonToKeep= [x for x in genesToKeep if x not in s]
print len(toRemove)
print len(genesToKeep)
print len(notcommonToKeep)
pathfiles=os.path.dirname(geneFile)
pathfiles=pathfiles+"/"
g_fp = HTSeq.FastaReader( genes )
removedparalogs=0
removedsize=0
totalgenes=0
rest=0
concatenatedFile=''
for contig in g_fp:
totalgenes+=1
name = contig.name+" "+contig.descr
name2= contig.name
if name2 not in toRemove and name2 in genesToKeep:
if int(len(contig.seq))>sizethresh:
namefile=contig.name
namefile=namefile.replace("|","_")
with open(pathfiles+namefile+".fasta", "wb") as f:
f.write(">1\n"+contig.seq+"\n")
rest+=1
concatenatedFile+=">"+namefile+"\n"+contig.seq+"\n"
else:
removedsize+=1
else:
removedparalogs+=1
print "%s genes are contained in other genes" % (g)
print "Removed %s same Locus genes" % str(removedparalogs)
print "Removed %s because of size " % str(removedsize)
print "%s Scheme genes " % str(rest)
print "total genes:" + str(totalgenes)
with open (pathfiles+"concatenated.fasta","wb") as f:
f.write (concatenatedFile)
def main():
print("Starting script at : " + time.strftime("%H:%M:%S-%d/%m/%Y"))
try:
input_file = sys.argv[1]
temppath = sys.argv[2]
except IndexError:
print "usage: list_pickle_obj"
argumentList = []
with open(input_file, 'rb') as f:
argumentList = pickle.load(f)
geneFile = argumentList[0]
genomesList = argumentList[1]
basepath = os.path.join(temppath, os.path.splitext(geneFile)[0])
if not os.path.exists(basepath):
os.makedirs(basepath)
gene_fp = HTSeq.FastaReader(geneFile)
alleleI = 0
resultsList = []
i = 0
perfectMatchIdAllele = []
perfectMatchIdAllele2 = []
allelescores = []
print("Getting BSR at : " + time.strftime("%H:%M:%S-%d/%m/%Y"))
geneScorePickle = os.path.abspath(geneFile) + '_bsr.txt'
#check if bsr as arealdy been calculated and recalculate it
if os.path.isfile(geneScorePickle):
alleleI, allelescores, alleleList = getBlastScoreRatios(
geneFile, basepath, False)
else:
alleleI, allelescores, alleleList = getBlastScoreRatios(
geneFile, basepath, True)
print("Finished BSR at : " + time.strftime("%H:%M:%S-%d/%m/%Y"))
genome = -1
genomeDict = {}
print("starting allele call at: " + time.strftime("%H:%M:%S-%d/%m/%Y"))
for genomeFile in genomesList:
print genomeFile
bestmatch = [
0, 0, False, '', 0
] #score, score ratio, perfectmatch, key name of the DNA sequence string, allele ID
currentGenomeDict = {}
currentCDSDict = {}
# load the translated CDS from the genome to a dictionary
filepath = os.path.join(
temppath,
str(os.path.basename(genomeFile)) + "_ORF_Protein.txt")
with open(filepath, 'rb') as f:
currentCDSDict = pickle.load(f)
#load the contig info of the genome to a dictionary
g_fp = HTSeq.FastaReader(genomeFile)
for contig in g_fp:
sequence = str(contig.seq)
genomeDict[contig.name] = sequence
currentGenomeDict = genomeDict
genome += 1
listOfCDS = currentCDSDict
genomeProteinfastaPath = os.path.join(
temppath, str(os.path.basename(genomeFile) + '_Protein.fasta'))
print("Blasting alleles on genome at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_out_file = os.path.join(
basepath, "blastdbs/" + os.path.basename(geneFile) + '_List.xml')
Gene_Blast_DB_name = os.path.join(
temppath,
str(os.path.basename(genomeFile)) + "/" +
str(os.path.basename(genomeFile)) + "_db")
proteinfastaPath = os.path.join(
basepath, str(os.path.basename(geneFile) + '_protein.fasta'))
#blast the genome CDS against the translated locus
cline = NcbiblastpCommandline(query=proteinfastaPath,
db=Gene_Blast_DB_name,
evalue=0.001,
out=blast_out_file,
outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, proteinfastaPath)
print("Blasted alleles on genome at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
alleleSizes = []
for allele in alleleList:
alleleSizes.append(len(allele))
biggestSizeAllele = 0
moda = max(set(alleleSizes), key=alleleSizes.count)
contador = Counter(alleleSizes).most_common()
if (contador[0])[1] == 1:
moda = alleleSizes[0]
try:
# iterate through the blast results
for blast_record in blast_records:
locationcontigs = []
for alignment in blast_record.alignments:
# select the best match
for match in alignment.hsps:
alleleMatchid = str(
blast_record.query_id).split("_")[1]
scoreRatio = float(match.score) / float(
allelescores[int(alleleMatchid) - 1])
cdsStrName = ((alignment.title).split(" "))[1]
DNAstr = listOfCDS[">" + cdsStrName]
AlleleDNAstr = alleleList[int(alleleMatchid) - 1]
if len(AlleleDNAstr) > biggestSizeAllele:
biggestSizeAllele = len(AlleleDNAstr)
compare = False
#compare the DNA match and the allele DNA sequence (protein sequences may be equal and DNA different)
if DNAstr == AlleleDNAstr is False:
try:
DNAstr = reverseComplement(DNAstr)
if DNAstr == AlleleDNAstr is False:
pass
else:
compare = True
except:
pass
else:
compare = True
if scoreRatio > 0.6:
locationcontigs.append(cdsStrName)
if "N" in DNAstr or "K" in DNAstr or "R" in DNAstr:
pass
elif (scoreRatio == 1 and bestmatch[2] is False
and compare is True):
bestmatch = [
match.score, scoreRatio, True, cdsStrName,
int(alleleMatchid), match,
len(AlleleDNAstr)
]
elif (scoreRatio == 1 and match.score > bestmatch[0]
and compare is True):
bestmatch = [
match.score, scoreRatio, True, cdsStrName,
int(alleleMatchid), match,
len(AlleleDNAstr)
]
elif (scoreRatio == 1 and bestmatch[2] is False
and compare is False):
bestmatch = [
match.score, scoreRatio, False, cdsStrName,
int(alleleMatchid), match,
len(AlleleDNAstr)
]
elif (scoreRatio == 1 and match.score > bestmatch[0]
and compare is False):
bestmatch = [
match.score, scoreRatio, False, cdsStrName,
int(alleleMatchid), match,
len(AlleleDNAstr)
]
elif (match.score > bestmatch[0] and scoreRatio > 0.6
and scoreRatio > bestmatch[1]
and bestmatch[2] is False):
bestmatch = [
match.score, scoreRatio, False, cdsStrName,
int(alleleMatchid), match,
len(AlleleDNAstr)
]
print("Classifying the match at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
#if no best match was found it's a Locus Not Found
if bestmatch[
0] == 0 or "N" in AlleleDNAstr or "K" in AlleleDNAstr or "R" in AlleleDNAstr:
###################
# LOCUS NOT FOUND #
###################
if bestmatch[0] == 0:
resultsList.append('LNF3:-1')
perfectMatchIdAllele.append('LNF')
perfectMatchIdAllele2.append('LNF')
print "Locus not found, no matches \n"
else:
resultsList.append('LNFN:-1')
perfectMatchIdAllele.append('LNF')
perfectMatchIdAllele2.append('LNF')
print "Locus has strange base (N, K or R) \n"
#if more than one BSR >0.6 in two different CDSs it's a Non Paralog Locus
elif len(list(set(locationcontigs))) > 1:
resultsList.append('NIPL')
perfectMatchIdAllele.append('NIPL')
perfectMatchIdAllele2.append('NIPL')
for elem in locationcontigs:
print elem
#in case the DNA match sequence equal to the DNA sequence of the comparing allele
elif bestmatch[2] is True:
contigname = bestmatch[3]
contigname = contigname.split("&")
matchLocation = contigname[2]
contigname = contigname[0]
print contigname
alleleStr = listOfCDS[">" + bestmatch[3]]
protSeq, alleleStr, Reversed = translateSeq(alleleStr)
#check for possible locus on tip
match = bestmatch[5]
matchLocation2 = matchLocation.split("-")
seq = currentGenomeDict[contigname]
bestMatchContigLen = len(seq)
rightmatchContig = bestMatchContigLen - int(matchLocation2[1])
leftmatchContig = int(matchLocation2[0])
if Reversed:
aux = rightmatchContig
rightmatchContig = leftmatchContig
leftmatchContig = aux
# get extra space to the right and left between the allele and match
possibleExtra = int(moda) - ((int(match.query_end) * 3) -
(int(match.query_start) * 3))
if possibleExtra < 0:
perfectMatchIdAllele.append(str(bestmatch[4]))
if not Reversed:
perfectMatchIdAllele2.append(
str(contigname) + "&" + str(matchLocation) + "&" +
"+")
else:
perfectMatchIdAllele2.append(
str(contigname) + "&" + str(matchLocation) + "&" +
"-")
resultsList.append('EXC:' + str(bestmatch[4]))
else:
rightmatchAllele = possibleExtra
leftmatchAllele = possibleExtra
if leftmatchContig < leftmatchAllele and rightmatchContig < rightmatchAllele:
resultsList.append('PLOTSC:-1')
perfectMatchIdAllele.append('PLOTSC')
perfectMatchIdAllele2.append('PLOTSC')
print match
print "contig extras (l,r)"
print leftmatchContig, rightmatchContig
print "allele extras (l,r)"
print leftmatchAllele, rightmatchAllele
print "Locus is possibly bigger than the contig \n"
elif leftmatchContig < leftmatchAllele:
resultsList.append('PLOT3:-1')
perfectMatchIdAllele.append('PLOT3')
perfectMatchIdAllele2.append('PLOT3')
print match
print "contig extras (l,r)"
print leftmatchContig, rightmatchContig
print "allele extras (l,r)"
print leftmatchAllele, rightmatchAllele
print "Locus is possibly on the 3' tip of the contig \n"
elif rightmatchContig < rightmatchAllele:
resultsList.append('PLOT5:-1')
perfectMatchIdAllele.append('PLOT5')
perfectMatchIdAllele2.append('PLOT5')
print match
print "contig extras (l,r)"
print leftmatchContig, rightmatchContig
print "allele extras (l,r)"
print leftmatchAllele, rightmatchAllele
print "Locus is possibly on the 5' tip of the contig \n"
else:
#if a perfect match was found
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
perfectMatchIdAllele.append(str(bestmatch[4]))
if not Reversed:
perfectMatchIdAllele2.append(
str(contigname) + "&" + str(matchLocation) +
"&" + "+")
else:
perfectMatchIdAllele2.append(
str(contigname) + "&" + str(matchLocation) +
"&" + "-")
resultsList.append('EXC:' + str(bestmatch[4]))
# if match with BSR >0.6 and not equal DNA sequences
else:
match = bestmatch[5]
geneLen = bestmatch[6]
contigname = bestmatch[3]
contigname = contigname.split("&")
matchLocation = contigname[2]
matchLocation = matchLocation.split("-")
contigname = contigname[0]
seq = currentGenomeDict[contigname]
bestMatchContigLen = len(seq)
alleleStr = listOfCDS[">" + bestmatch[3]]
protSeq, alleleStr, Reversed = translateSeq(alleleStr)
rightmatchContig = bestMatchContigLen - int(matchLocation[1])
leftmatchContig = int(matchLocation[0])
if Reversed:
aux = rightmatchContig
rightmatchContig = leftmatchContig
leftmatchContig = aux
print rightmatchContig, leftmatchContig
# get extra space to the right and left between the allele and match and check if it's still inside the contig
rightmatchAllele = geneLen - ((int(match.query_end) + 1) * 3)
leftmatchAllele = ((int(match.query_start) - 1) * 3)
###########################
# LOCUS ON THE CONTIG TIP #
###########################
if leftmatchContig < leftmatchAllele and rightmatchContig < rightmatchAllele:
resultsList.append('LOTSC:-1')
perfectMatchIdAllele.append('LOTSC')
perfectMatchIdAllele2.append('LOTSC')
print match
print contigname
print geneFile
print leftmatchAllele, rightmatchAllele
print "Locus is bigger than the contig \n"
elif leftmatchContig < leftmatchAllele:
resultsList.append('LOT3:-1')
perfectMatchIdAllele.append('LOT3')
perfectMatchIdAllele2.append('LOT3')
print match
print contigname
print geneFile
print leftmatchAllele, rightmatchAllele
print "Locus is on the 3' tip of the contig \n"
elif rightmatchContig < rightmatchAllele:
resultsList.append('LOT5:-1')
perfectMatchIdAllele.append('LOT5')
perfectMatchIdAllele2.append('LOT5')
print match
print contigname
print geneFile
print leftmatchAllele, rightmatchAllele
print "Locus is on the 5' tip of the contig \n"
elif len(alleleStr) > moda + (moda * 0.2):
print moda
print alleleStr
resultsList.append('ALM')
perfectMatchIdAllele.append('ALM')
perfectMatchIdAllele2.append('ALM')
elif len(alleleStr) < moda - (moda * 0.2):
print moda
print alleleStr
resultsList.append('ASM')
perfectMatchIdAllele.append('ASM')
perfectMatchIdAllele2.append('ASM')
else:
#######################
# ADD INFERRED ALLELE # # a new allele
#######################
tagAux = 'INF'
perfectMatchIdAllele.append(tagAux + "-" +
str(alleleI + 1))
if not Reversed:
perfectMatchIdAllele2.append(
str(contigname) + "&" + str(matchLocation[0]) +
"-" + str(matchLocation[1]) + "&" + "+")
else:
perfectMatchIdAllele2.append(
str(contigname) + "&" + str(matchLocation[0]) +
"-" + str(matchLocation[1]) + "&" + "-")
print "New allele! Adding allele " + tagAux + str(
alleleI + 1) + " to the database\n"
resultsList.append(tagAux + str(alleleI + 1))
# --- add the new allele to the gene fasta --- #
appendAllele = '>allele_' + str(
alleleI + 1) + '_' + tagAux[:-1] + "_" + str(
os.path.basename(genomesList[genome])) + '\n'
fG = open(geneFile, 'a')
fG.write(appendAllele)
fG.write(alleleStr + '\n')
fG.close()
fG = open(
os.path.join(
basepath,
str(
os.path.basename(geneFile) +
'_protein2.fasta')), 'w')
fG.write('>' + str(alleleI + 1) + '\n' + str(protSeq) +
'\n')
fG.close()
fG = open(
os.path.join(
basepath,
str(os.path.basename(geneFile) +
'_protein.fasta')), 'a')
fG.write('>' + str(alleleI + 1) + '\n' + str(protSeq) +
'\n')
fG.close()
match = bestmatch[5]
# --- remake blast DB and recalculate the BSR for the locus --- #
alleleList.append(alleleStr)
print os.path.join(
basepath,
str(os.path.basename(geneFile) + '_protein.fasta'))
genefile2 = os.path.join(
basepath,
str(os.path.basename(geneFile) + '_protein2.fasta'))
Gene_Blast_DB_name2 = Create_Blastdb(genefile2, 1, True)
print("Re-calculating BSR at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
alleleI, allelescores, alleleList = reDogetBlastScoreRatios(
genefile2, basepath, alleleI, allelescores,
Gene_Blast_DB_name2, alleleList, geneScorePickle)
print "allele id " + str(alleleI)
print("Done Re-calculating BSR at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
except Exception as e:
print "some error occurred"
print e
print 'Error on line {}'.format(sys.exc_info()[-1].tb_lineno)
perfectMatchIdAllele2.append("ERROR")
perfectMatchIdAllele.append("ERROR")
resultsList.append('ERROR')
final = (resultsList, perfectMatchIdAllele)
print("Finished allele calling at : " + time.strftime("%H:%M:%S-%d/%m/%Y"))
filepath = os.path.join(temppath,
os.path.basename(geneFile) + "_result.txt")
filepath2 = os.path.join(temppath,
os.path.basename(geneFile) + "_result2.txt")
with open(filepath, 'wb') as f:
pickle.dump(final, f)
with open(filepath2, 'wb') as f:
pickle.dump(perfectMatchIdAllele2, f)
shutil.rmtree(basepath)
return True
def getBlastScoreRatios(orgName,allelescores,cdsDict,prodigalPath):
openPresults = prodigalPath
Presults=open(openPresults, 'r')
linesP = Presults.readlines()
lastlineP=len(linesP)
alleleProt=''
proteome=""
countP=0
countCDS=0
if isContig=="no":
CreateProteome(nameOrg)
else:
queryCDS = CreateProteomeContig(nameOrg,cdsDict)
cline = NcbiblastpCommandline(query=pathRef+nameOrg+'proteome.fasta', db=name, out=blast_out_file, outfmt=5, num_alignments=7000, num_descriptions=7000)
#print cline
allelescore=0
blast_records = runBlastParser(cline,blast_out_file, alleleProt)
os.remove(pathRef+nameOrg+'proteome.fasta')
blastScoreRatio=0
countRecords=0
bestMatches={}
BestMatchResults= []
length=[]
alignment_posStart=[]
query_length=[]
for blast_record in blast_records:
found=False
countRecords+=1
for alignment in blast_record.alignments:
if found is False:
#print blast_record.query, alignment.hit_def
scoreToUse=0
for match in alignment.hsps:
if len(blast_record.alignments)==0:
countResults=countResults
else:
blastScoreRatio = float(match.score) / float(allelescores[str(alignment.hit_def)])
#or re.search("ENA|",alignment.title)
#print alignment.title
try:
geneName=alignment.title.split("|")[5]
except IndexError:
geneName=alignment.title.split("|")[2]
#print geneName
#products.append(alignment.title.split("|")[6].split("[")[0])
#if hsp.expect < 0.001 and 100 <= hsp.align_length:
if geneName.strip() not in BestMatchResults and blastScoreRatio>0.6:
BestMatchResults.append(genomeDB+"..."+str(geneName).strip())
length.append(str(match.align_length-1))
#score.append(str(Score))
alignment_posStart.append(str(match.query_start))
query_length.append(str(len(match.query)))
break
else:
break
#print str(blast_record.query)
bestMatches[str(countRecords)] = [BestMatchResults,length,alignment_posStart,query_length,str(blast_record.query)]
BestMatchResults= []
length=[]
alignment_posStart=[]
query_length=[]
print countRecords
#fG = open( pathRef+'AllAlleles.fasta', 'a' )
#for i in ToNewAllele:
#print i
#fG.write(i)
#fG.close()
#Create_Blastdb(pathRef+'allAllelesAA.fasta',1,True)
#print matchR
#print allelescores
#return alleleI,allelescores,Gene_Blast_DB_name
#print alleleI
#print len(allelescores)
#print countT
return bestMatches,queryCDS
def getBlastScoreRatios(genefile, basepath, doAll):
gene_fp = HTSeq.FastaReader(genefile)
alleleI = 0
allelescores = []
alleleProt = ''
alleleAllProt = ''
alleleList = []
for allele in gene_fp: #new db for each allele to blast it against himself
alleleI += 1
genome = -1
alleleList.append(allele.seq)
translatedSequence, x, y = translateSeq(allele.seq)
if translatedSequence == '':
pass
else:
alleleProt = ">" + str(alleleI) + "\n" + str(translatedSequence +
"\n")
alleleAllProt += ">" + str(alleleI) + "\n" + str(
translatedSequence + "\n")
proteinfastaPath = os.path.join(
basepath, str(os.path.basename(genefile) + '_protein2.fasta'))
with open(proteinfastaPath, "wb") as f:
f.write(alleleProt)
Gene_Blast_DB_name = Create_Blastdb(proteinfastaPath, 1, True)
if doAll:
blast_out_file = os.path.join(basepath, 'blastdbs/temp.xml')
print("Starting Blast alleles at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
# --- get BLAST score ratio --- #
cline = NcbiblastpCommandline(query=proteinfastaPath,
db=Gene_Blast_DB_name,
evalue=0.001,
out=blast_out_file,
outfmt=5)
allelescore = 0
blast_records = runBlastParser(cline, blast_out_file,
alleleProt)
print("Blasted alleles at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
for blast_record in blast_records:
for alignment in blast_record.alignments:
for match in alignment.hsps:
allelescores.append(int(match.score))
geneScorePickle = os.path.abspath(genefile) + '_bsr.txt'
print "________"
var = [alleleI, allelescores]
with open(geneScorePickle, 'wb') as f:
pickle.dump(var, f)
else:
geneScorePickle = os.path.abspath(genefile) + '_bsr.txt'
with open(geneScorePickle, 'rb') as f:
var = pickle.load(f)
allelescores = var[1]
proteinfastaPath = os.path.join(
basepath, str(os.path.basename(genefile) + '_protein.fasta'))
with open(proteinfastaPath, "wb") as f:
f.write(alleleAllProt)
return int(alleleI), allelescores, alleleList
def getBlastScoreRatios(genefile,basepath,doAll):
gene_fp = HTSeq.FastaReader(genefile)
alleleI=0
allelescores=[]
alleleProt=''
alleleAllProt=''
alleleList=[]
for allele in gene_fp: #new db for each allele to blast it against himself
alleleI+=1
genome=-1
alleleList.append(allele.seq)
translatedSequence,x,y=translateSeq(allele.seq)
if translatedSequence =='':
pass
else:
alleleProt=">"+str(alleleI)+"\n"+str(translatedSequence+"\n")
alleleAllProt+=">"+str(alleleI)+"\n"+str(translatedSequence+"\n")
proteinfastaPath=os.path.join(basepath,str(os.path.basename(genefile)+'_protein2.fasta'))
with open(proteinfastaPath, "wb") as f:
f.write(alleleProt)
Gene_Blast_DB_name = Create_Blastdb( proteinfastaPath, 1, True )
if doAll:
blast_out_file = os.path.join(basepath,'blastdbs/temp.xml')
print ("Starting Blast alleles at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
# --- get BLAST score ratio --- #
cline = NcbiblastpCommandline(query=proteinfastaPath, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
allelescore=0
blast_records = runBlastParser(cline,blast_out_file, alleleProt)
print ("Blasted alleles at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
for blast_record in blast_records:
for alignment in blast_record.alignments:
for match in alignment.hsps:
allelescores.append(int(match.score))
geneScorePickle=os.path.abspath(genefile)+'_bsr.txt'
print "________"
var=[alleleI,allelescores]
with open(geneScorePickle,'wb') as f:
pickle.dump(var, f)
else:
geneScorePickle=os.path.abspath(genefile)+'_bsr.txt'
with open(geneScorePickle,'rb') as f:
var = pickle.load(f)
allelescores=var[1]
proteinfastaPath=os.path.join(basepath,str(os.path.basename(genefile)+'_protein.fasta'))
with open(proteinfastaPath, "wb") as f:
f.write(alleleAllProt)
return int(alleleI),allelescores,alleleList
def main():
print ("Starting script at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
try:
input_file = sys.argv[1]
temppath = sys.argv[2]
except IndexError:
print "usage: list_pickle_obj"
argumentList=[]
with open(input_file,'rb') as f:
argumentList = pickle.load(f)
geneFile = argumentList[0]
genomesList = argumentList[1]
basepath=os.path.join(temppath,os.path.splitext(geneFile)[0])
if not os.path.exists(basepath):
os.makedirs(basepath)
gene_fp = HTSeq.FastaReader(geneFile)
alleleI = 0
resultsList = []
i = 0
perfectMatchIdAllele=[]
perfectMatchIdAllele2=[]
allelescores=[]
print ("Getting BSR at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
geneScorePickle=os.path.abspath(geneFile)+'_bsr.txt'
#check if bsr as arealdy been calculated and recalculate it
if os.path.isfile(geneScorePickle) :
alleleI,allelescores,alleleList=getBlastScoreRatios(geneFile,basepath,False)
else:
alleleI,allelescores,alleleList=getBlastScoreRatios(geneFile,basepath,True)
print ("Finished BSR at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
genome=-1
genomeDict = {}
print ("starting allele call at: "+time.strftime("%H:%M:%S-%d/%m/%Y"))
for genomeFile in genomesList:
print genomeFile
bestmatch=[0,0,False,'',0] #score, score ratio, perfectmatch, key name of the DNA sequence string, allele ID
currentGenomeDict={}
currentCDSDict={}
# load the translated CDS from the genome to a dictionary
filepath=os.path.join(temppath,str(os.path.basename(genomeFile))+"_ORF_Protein.txt")
with open(filepath,'rb') as f:
currentCDSDict = pickle.load(f)
#load the contig info of the genome to a dictionary
g_fp = HTSeq.FastaReader( genomeFile )
for contig in g_fp:
sequence=str(contig.seq)
genomeDict[ contig.name ] = sequence
currentGenomeDict = genomeDict
genome+=1
listOfCDS=currentCDSDict
genomeProteinfastaPath=os.path.join(temppath,str(os.path.basename(genomeFile)+'_Protein.fasta'))
print ("Blasting alleles on genome at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_out_file = os.path.join(basepath,"blastdbs/"+os.path.basename(geneFile)+ '_List.xml')
Gene_Blast_DB_name = os.path.join(temppath,str(os.path.basename(genomeFile))+"/"+str(os.path.basename(genomeFile))+"_db")
proteinfastaPath=os.path.join(basepath,str(os.path.basename(geneFile)+'_protein.fasta'))
#blast the genome CDS against the translated locus
cline = NcbiblastpCommandline(query=proteinfastaPath, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, proteinfastaPath)
print ("Blasted alleles on genome at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
alleleSizes=[]
for allele in alleleList:
alleleSizes.append(len(allele))
biggestSizeAllele=0
moda=max(set(alleleSizes), key=alleleSizes.count)
contador= Counter(alleleSizes).most_common()
if (contador[0])[1] ==1:
moda= alleleSizes[0]
try:
# iterate through the blast results
for blast_record in blast_records:
locationcontigs=[]
for alignment in blast_record.alignments:
# select the best match
for match in alignment.hsps:
alleleMatchid=str(blast_record.query_id).split("_")[1]
scoreRatio=float(match.score)/float(allelescores[int(alleleMatchid)-1])
cdsStrName=((alignment.title).split(" "))[1]
DNAstr=listOfCDS[">"+cdsStrName]
AlleleDNAstr=alleleList[int(alleleMatchid)-1]
if len(AlleleDNAstr)>biggestSizeAllele:
biggestSizeAllele=len(AlleleDNAstr)
compare=False
#compare the DNA match and the allele DNA sequence (protein sequences may be equal and DNA different)
if DNAstr==AlleleDNAstr is False:
try:
DNAstr=reverseComplement(DNAstr)
if DNAstr==AlleleDNAstr is False:
pass
else:
compare=True
except:
pass
else:
compare=True
if scoreRatio>0.6:
locationcontigs.append(cdsStrName)
if "N" in DNAstr or "K" in DNAstr or "R" in DNAstr:
pass
elif(scoreRatio == 1 and bestmatch[2] is False and compare is True):
bestmatch=[match.score,scoreRatio,True,cdsStrName,int(alleleMatchid),match,len(AlleleDNAstr)]
elif(scoreRatio == 1 and match.score>bestmatch[0] and compare is True):
bestmatch=[match.score,scoreRatio,True,cdsStrName,int(alleleMatchid),match,len(AlleleDNAstr)]
elif(scoreRatio == 1 and bestmatch[2] is False and compare is False):
bestmatch=[match.score,scoreRatio,False,cdsStrName,int(alleleMatchid),match,len(AlleleDNAstr)]
elif(scoreRatio == 1 and match.score>bestmatch[0] and compare is False):
bestmatch=[match.score,scoreRatio,False,cdsStrName,int(alleleMatchid),match,len(AlleleDNAstr)]
elif(match.score>bestmatch[0] and scoreRatio>0.6 and scoreRatio>bestmatch[1] and bestmatch[2] is False):
bestmatch=[match.score,scoreRatio,False,cdsStrName,int(alleleMatchid),match,len(AlleleDNAstr)]
print ("Classifying the match at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
#if no best match was found it's a Locus Not Found
if bestmatch[0]==0 or "N" in AlleleDNAstr or "K" in AlleleDNAstr or "R" in AlleleDNAstr :
###################
# LOCUS NOT FOUND #
###################
if bestmatch[0]==0:
resultsList.append('LNF3:-1')
perfectMatchIdAllele.append('LNF')
perfectMatchIdAllele2.append('LNF')
print "Locus not found, no matches \n"
else:
resultsList.append('LNFN:-1')
perfectMatchIdAllele.append('LNF')
perfectMatchIdAllele2.append('LNF')
print "Locus has strange base (N, K or R) \n"
#if more than one BSR >0.6 in two different CDSs it's a Non Paralog Locus
elif len(list(set(locationcontigs)))>1:
resultsList.append('NIPL')
perfectMatchIdAllele.append('NIPL')
perfectMatchIdAllele2.append('NIPL')
for elem in locationcontigs:
print elem
#in case the DNA match sequence equal to the DNA sequence of the comparing allele
elif bestmatch[2] is True:
contigname=bestmatch[3]
contigname=contigname.split("&")
matchLocation=contigname[2]
contigname=contigname[0]
print contigname
alleleStr=listOfCDS[">"+bestmatch[3]]
protSeq,alleleStr,Reversed=translateSeq(alleleStr)
#check for possible locus on tip
match=bestmatch[5]
matchLocation2=matchLocation.split("-")
seq=currentGenomeDict[ contigname ]
bestMatchContigLen=len(seq)
rightmatchContig=bestMatchContigLen-int(matchLocation2[1])
leftmatchContig=int(matchLocation2[0])
if Reversed:
aux=rightmatchContig
rightmatchContig=leftmatchContig
leftmatchContig=aux
# get extra space to the right and left between the allele and match
possibleExtra=int(moda)-((int(match.query_end)*3)-(int(match.query_start)*3))
if possibleExtra<0:
perfectMatchIdAllele.append(str(bestmatch[4]))
if not Reversed:
perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation)+"&"+"+")
else:
perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation)+"&"+"-")
resultsList.append('EXC:' + str(bestmatch[4]) )
else:
rightmatchAllele=possibleExtra
leftmatchAllele=possibleExtra
if leftmatchContig<leftmatchAllele and rightmatchContig < rightmatchAllele:
resultsList.append('PLOTSC:-1')
perfectMatchIdAllele.append('PLOTSC')
perfectMatchIdAllele2.append('PLOTSC')
print match
print "contig extras (l,r)"
print leftmatchContig,rightmatchContig
print "allele extras (l,r)"
print leftmatchAllele,rightmatchAllele
print "Locus is possibly bigger than the contig \n"
elif leftmatchContig<leftmatchAllele:
resultsList.append('PLOT3:-1')
perfectMatchIdAllele.append('PLOT3')
perfectMatchIdAllele2.append('PLOT3')
print match
print "contig extras (l,r)"
print leftmatchContig,rightmatchContig
print "allele extras (l,r)"
print leftmatchAllele,rightmatchAllele
print "Locus is possibly on the 3' tip of the contig \n"
elif rightmatchContig < rightmatchAllele:
resultsList.append('PLOT5:-1')
perfectMatchIdAllele.append('PLOT5')
perfectMatchIdAllele2.append('PLOT5')
print match
print "contig extras (l,r)"
print leftmatchContig,rightmatchContig
print "allele extras (l,r)"
print leftmatchAllele,rightmatchAllele
print "Locus is possibly on the 5' tip of the contig \n"
else:
#if a perfect match was found
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
perfectMatchIdAllele.append(str(bestmatch[4]))
if not Reversed:
perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation)+"&"+"+")
else:
perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation)+"&"+"-")
resultsList.append('EXC:' + str(bestmatch[4]) )
# if match with BSR >0.6 and not equal DNA sequences
else:
match=bestmatch[5]
geneLen=bestmatch[6]
contigname=bestmatch[3]
contigname=contigname.split("&")
matchLocation=contigname[2]
matchLocation=matchLocation.split("-")
contigname=contigname[0]
seq=currentGenomeDict[ contigname ]
bestMatchContigLen=len(seq)
alleleStr=listOfCDS[">"+bestmatch[3]]
protSeq,alleleStr,Reversed=translateSeq(alleleStr)
rightmatchContig=bestMatchContigLen-int(matchLocation[1])
leftmatchContig=int(matchLocation[0])
if Reversed:
aux=rightmatchContig
rightmatchContig=leftmatchContig
leftmatchContig=aux
print rightmatchContig,leftmatchContig
# get extra space to the right and left between the allele and match and check if it's still inside the contig
rightmatchAllele=geneLen-((int(match.query_end)+1)*3)
leftmatchAllele=((int(match.query_start)-1)*3)
###########################
# LOCUS ON THE CONTIG TIP #
###########################
if leftmatchContig<leftmatchAllele and rightmatchContig < rightmatchAllele:
resultsList.append('LOTSC:-1')
perfectMatchIdAllele.append('LOTSC')
perfectMatchIdAllele2.append('LOTSC')
print match
print contigname
print geneFile
print leftmatchAllele,rightmatchAllele
print "Locus is bigger than the contig \n"
elif leftmatchContig<leftmatchAllele:
resultsList.append('LOT3:-1')
perfectMatchIdAllele.append('LOT3')
perfectMatchIdAllele2.append('LOT3')
print match
print contigname
print geneFile
print leftmatchAllele,rightmatchAllele
print "Locus is on the 3' tip of the contig \n"
elif rightmatchContig < rightmatchAllele:
resultsList.append('LOT5:-1')
perfectMatchIdAllele.append('LOT5')
perfectMatchIdAllele2.append('LOT5')
print match
print contigname
print geneFile
print leftmatchAllele,rightmatchAllele
print "Locus is on the 5' tip of the contig \n"
elif len(alleleStr) > moda+(moda*0.2) :
print moda
print alleleStr
resultsList.append('ALM')
perfectMatchIdAllele.append('ALM')
perfectMatchIdAllele2.append('ALM')
elif len(alleleStr) < moda-(moda*0.2):
print moda
print alleleStr
resultsList.append('ASM')
perfectMatchIdAllele.append('ASM')
perfectMatchIdAllele2.append('ASM')
else:
#######################
# ADD INFERRED ALLELE # # a new allele
#######################
tagAux='INF'
perfectMatchIdAllele.append( tagAux +"-"+str(alleleI+1))
if not Reversed:
perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"+")
else:
perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"-")
print "New allele! Adding allele "+ tagAux + str(alleleI+1) +" to the database\n"
resultsList.append( tagAux + str(alleleI+1) )
# --- add the new allele to the gene fasta --- #
appendAllele='>allele_' + str(alleleI+1) + '_' + tagAux[:-1] +"_" + str(os.path.basename(genomesList[genome])) + '\n'
fG = open( geneFile, 'a' )
fG.write(appendAllele)
fG.write( alleleStr + '\n')
fG.close()
fG = open( os.path.join(basepath,str(os.path.basename(geneFile)+'_protein2.fasta')), 'w' )
fG.write('>'+str(alleleI+1)+'\n'+str(protSeq) + '\n')
fG.close()
fG = open( os.path.join(basepath,str(os.path.basename(geneFile)+'_protein.fasta')), 'a' )
fG.write('>'+str(alleleI+1)+'\n'+str(protSeq) + '\n')
fG.close()
match=bestmatch[5]
# --- remake blast DB and recalculate the BSR for the locus --- #
alleleList.append(alleleStr)
print os.path.join(basepath,str(os.path.basename(geneFile)+'_protein.fasta'))
genefile2= os.path.join(basepath,str(os.path.basename(geneFile)+'_protein2.fasta'))
Gene_Blast_DB_name2 = Create_Blastdb( genefile2, 1, True )
print ("Re-calculating BSR at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
alleleI,allelescores,alleleList=reDogetBlastScoreRatios(genefile2,basepath,alleleI,allelescores,Gene_Blast_DB_name2,alleleList,geneScorePickle)
print "allele id " + str(alleleI)
print ("Done Re-calculating BSR at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
except Exception as e:
print "some error occurred"
print e
print 'Error on line {}'.format(sys.exc_info()[-1].tb_lineno)
perfectMatchIdAllele2.append("ERROR")
perfectMatchIdAllele.append("ERROR")
resultsList.append('ERROR')
final = (resultsList,perfectMatchIdAllele)
print ("Finished allele calling at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
filepath=os.path.join(temppath , os.path.basename(geneFile)+"_result.txt")
filepath2=os.path.join(temppath , os.path.basename(geneFile)+"_result2.txt")
with open(filepath, 'wb') as f:
pickle.dump(final, f)
with open(filepath2, 'wb') as f:
pickle.dump(perfectMatchIdAllele2, f)
shutil.rmtree(basepath)
return True
def main():
parser = argparse.ArgumentParser(description="Given two list of genes, creates a folder with paired files when located on the same locus")
parser.add_argument('-i', nargs='?', type=str, help='1st list of genes files to compare', required=True)
parser.add_argument('-g', nargs='?', type=str, help='2nd list of genes files to compare', required=True)
args = parser.parse_args()
geneFiles1 = args.i
geneFiles2 = args.g
name1="concat1.fasta"
name2="concat2.fasta"
concat_genes(geneFiles1, name1)
concat_genes(geneFiles2, name2)
#orderedAlleleNames=[]
geneDict={}
gene_fp = HTSeq.FastaReader(name1)
alleleI=0
for allele in gene_fp:
#if allele.seq in geneDict:
# print "\nWARNING: this file contains a repeated allele, it should be checked. Ignoring it now!\n", geneFile
#else:
#orderedAlleleNames.append(allele.name)
geneDict[ allele.seq ] = allele.name
alleleI += 1
gene_fp = HTSeq.FastaReader(name1)
geneFile = os.path.abspath( name1 )
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 , False)
geneF = os.path.splitext( geneFile )[0]
blast_out_file = geneF + '.xml'
# list of results - the output of the function
resultsList = []
# ------------------------------ RUNNING BLAST ------------------------------ #
cline = NcbiblastnCommandline(query=name2, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, name2)
samelocus=0
alreadyUsed=[]
nomatch=0
small=0
if not os.path.exists("./sameLocus"):
os.makedirs("./sameLocus")
LocusID=0
for blast_record in blast_records:
try:
alignment=blast_record.alignments[1]
#print blast_record.query
#print alignment.num_alignments
try:
#print alleleLength, alignment.length
i=0
align=blast_record.alignments[i]
while i<len(blast_record.alignments):
if align.hit_def:
result,allelename2,LocusID,alreadyUsed=alignHasGoodMatch(align,geneDict,LocusID,blast_record,alreadyUsed)
if result>0 and allelename2:
samelocus+=result
i+=999
else:
small+=1
i+=999
alreadyUsed.append(allelename2)
elif allelename :
#alreadyUsed.append(allelename)
result,allelename,LocusID,alreadyUsed=alignHasGoodMatch(align,geneDict,LocusID,blast_record,alreadyUsed)
if result>0:
samelocus+=result
i+=999
else:
small+=1
i+=999
#alreadyUsed.append(allelename2)
else :
nomatch+=1
#print align.length, alleleleng
i+=1
except Exception as e:
print e
#print "lkjh"
pass
except:
try:
alignment=blast_record.alignments[0]
#print blast_record.query
result,allelename,LocusID,alreadyUsed=alignHasGoodMatch(alignment,geneDict,LocusID,blast_record,alreadyUsed)
if result>0 and allelename:
samelocus+=result
else :
small+=1
#alreadyUsed.append(allelename)
#alreadyUsed.append(alignment.hit_def)
except:
nomatch+=1
print "%s are within same locus, %s had no match and %s had a bigger than 0.2 ratio size difference or less than 0.8 similarity ratio" % (samelocus,nomatch, small)
os.remove(name1)
os.remove(name2)
shutil.rmtree('./blastdbs')
def main():
try:
input_file = sys.argv[1]
temppath = sys.argv[2]
except IndexError:
print "usage: list_pickle_obj"
argumentList=[]
with open(input_file,'rb') as f:
argumentList = pickle.load(f)
geneFile = argumentList[0]
genomesList = argumentList[1]
#listOfCDSDicts = argumentList[2]
basepath=temppath
gene_fp = HTSeq.FastaReader(geneFile)
geneDict = {}
alleleI = 1
inverted=False
orderedAlleleNames=[]
biggestAllelelen=0
smallestAllelelen=99999
for allele in gene_fp:
if allele.seq in geneDict:
print "\nWARNING: this file contains a repeated allele, it should be checked. Ignoring it now!\n", geneFile
else:
if len(allele.seq)>biggestAllelelen:
biggestAllelelen=len(allele.seq)
if len(allele.seq)<smallestAllelelen:
smallestAllelelen=len(allele.seq)
orderedAlleleNames.append(allele.name)
geneDict[ allele.seq ] = alleleI
alleleI += 1
#print geneDict
#print orderedAlleleNames
# --- make 1st blast DB --- #
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1, False )
geneF = os.path.basename(geneFile)
blast_out_file = os.path.dirname(geneFile)+"/blastdbs/"+geneF + '.xml'
# list of results - the output of the function
resultsList = []
i = 0
perfectMatchIdAllele=[]
genomeDict = {}
for genomeFile in genomesList:
#currentCDSDict = listOfCDSDicts[i]
filepath=os.path.join(basepath,str(os.path.basename(genomeFile))+"_ORF.txt")
with open(filepath,'rb') as f:
currentCDSDict = pickle.load(f)
g_fp = HTSeq.FastaReader( genomeFile )
for contig in g_fp:
sequence=str(contig.seq)
genomeDict[ contig.name ] = sequence
currentGenomeDict = genomeDict
#print genomeFile
#print resultsList
#print geneDict
#print orderedAlleleNames
i+=1 # it has to be incremented here
if genomeFile[-1] == '\n':
genomeFile = genomeFile[:-1]
# ------------------------------ RUNNING BLAST ------------------------------ #
#print Gene_Blast_DB_name
#cline = NcbiblastnCommandline(query=genomeFile, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
cline = NcbiblastnCommandline(query=genomeFile, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
#print cline
blast_records = runBlastParser(cline, blast_out_file, genomeFile)
print ("Finished Blast at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
# ------ DETERMINING BEST MATCH ------ #
# bestMatch = ['rec.query','hsp', lenRatio]
bestMatch = ['','', 0]
bestMatchContig=''
bestMatchContigLen=''
bestalignlen=0
perfectMatch=False
bmAlleleLen2=0
bmAllele=''
#noAlignment=False
for blast_record in blast_records:
# --- the LNF cases are now called outside de loop --- #
#print blast_record
if perfectMatch==True:
break
try:
#print blast_record.alignments
hspC = blast_record.alignments[0]
if bestMatch[0] == '' and bestMatch[1] == '':
bestMatch[0] = blast_record.query
bestMatch[1] = hspC
except IndexError:
continue
# --- the contig tag is used in the progigal function --- #
contigTag = blast_record.query
# --- brute force parsing of the contig tag - better solution is advisable --- #
j=0
for l in contigTag:
if l == ' ':
break
j+=1
contigTag = contigTag[:j]
contigLen = blast_record.query_letters
#print blast_record.query_id
# --- iterating over all the results to determine the best match --- #
for alignment in blast_record.alignments:
index=orderedAlleleNames.index(alignment.hit_def)
#print alignment.hit_def
for k, v in geneDict.iteritems():
if v == index+1:
bmAlleleLen2= len(k)
if perfectMatch:
break
for match in alignment.hsps:
#print match
scoreRatio = float(match.score) / float(bmAlleleLen2)
#print alignment.hit_def
#print match.identities
#print bmAlleleLen2
#print
#print match.identities
#print len(match.sbjct)
#if #identities is the same as the length of the allele and it has no gaps or N's
if (int(match.identities)==int(bmAlleleLen2) and int(match.identities)==int(len(match.sbjct)) and "N" not in match.query and "K" not in match.query and "R" not in match.query ):
index=orderedAlleleNames.index(alignment.hit_def)
for seq, alleleid in geneDict.iteritems():
if alleleid == index+1:
bmAllele=seq
break
bmAlleleLen= len(bmAllele)
lenratio=float(len(match.query))/float(bmAlleleLen)
bestMatch = [blast_record.query, match, scoreRatio, alignment.hit_def,lenratio,bmAlleleLen]
bestMatchContig=contigTag
perfectMatch=True
index=orderedAlleleNames.index(alignment.hit_def)
bmAlleleLen= len(geneDict.keys()[index])
break
#choose the match with the best score ratio score/length of allele
elif scoreRatio > bestMatch[2]:
index=orderedAlleleNames.index(alignment.hit_def)
for seq, alleleid in geneDict.iteritems():
if alleleid == index+1:
bmAllele=seq
break
bmAlleleLen= len(bmAllele)
lenratio=float(len(match.query))/float(bmAlleleLen)
bestMatch = [blast_record.query, match, scoreRatio, alignment.hit_def,lenratio,bmAlleleLen]
bestMatchContig=contigTag
bestMatchContigLen=blast_record.query_letters
if match.sbjct_start > match.sbjct_end:
inverted=True
#print match.query
bestalignlen=alignment.length
#print match
#print bmAlleleLen, bestMatchContig
if perfectMatch==True:
break
# ---------- ALLELE CALLING AFTER DETERMINING BEST MATCH ---------- #
print ("Finished choosing best match at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
try:
#print bestMatch[0]
match = bestMatch[1]
#print match.query
geneLen = bestMatch[5]
alleleStr = match.query
nIdentities = match.identities
idPercent = float(nIdentities) / float(geneLen)
scoreRatio = bestMatch[2]
lenRatio = bestMatch[4]
#print perfectMatch
#print "\nContig best/exact match is :"
#print bestMatchContig +"\n"
except:
#if no best match was found
###################
# LOCUS NOT FOUND #
###################
resultsList.append('LNF3:-1') # append result to the list of results
perfectMatchIdAllele.append('LNF')
printinfo(genomeFile,geneFile)
print "Locus not found, no matches \n"
continue
if perfectMatch is True:
#if a perfect match was found
if match.sbjct_start > match.sbjct_end: #reverse the order if needed
alleleStr = reverseComplement(alleleStr)
alleleNumber = geneDict[ alleleStr ]
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
if "_" in bestMatch[3]:
a=bestMatch[3].split("_")
perfectMatchIdAllele.append(a[1])
else:
perfectMatchIdAllele.append(bestMatch[3])
resultsList.append('EXC:' + str(alleleNumber) )
elif bestMatch[0] != '' and perfectMatch is not True:
#if a best match was found but it's not an exact match
###########################
# LOCUS ON THE CONTIG TIP #
###########################
if match.query_start ==1 and len(match.query) < geneLen:
resultsList.append('LOT5:-1')
perfectMatchIdAllele.append('LOT5')
printinfo(genomeFile,geneFile)
print "Locus is on the 5' tip of the contig \n"
elif match.query_end == bestMatchContigLen and len(match.query) < geneLen:
resultsList.append('LOT3:-1')
perfectMatchIdAllele.append('LOT3')
printinfo(genomeFile,geneFile)
print "Locus is on the 3' tip of the contig \n"
elif bestMatchContigLen <= geneLen:
resultsList.append('LOTSC:-1')
perfectMatchIdAllele.append('LOTSC')
printinfo(genomeFile,geneFile)
#print match.query_start
print "Locus is bigger than the contig \n"
elif 'N' in alleleStr:
#TODO gravar para ficheiro
#####################
# ALLELE NOT FOUND # # N base found!
#####################
geneFile2= os.path.splitext(geneFile)[0] + "LNFN.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+"\n")
f.write((alleleStr) +"\n")
resultsList.append('LNFN:-1')
perfectMatchIdAllele.append('LNFN')
printinfo(genomeFile,geneFile)
print "LNFN, contains N bases! \n"
else:
# ------------------------------------------------------------------------------------------------------- #
# #
# USING PRODIGAL TO TRY TO EXTEND CDS #
# #
# ------------------------------------------------------------------------------------------------------- #
CDSType=''
sizeratio=0.2
ORFFoundInMatch, strCDS, CDSType = extendCDS(bestMatchContig, currentCDSDict, match.query_start, match.query_end, currentGenomeDict, biggestAllelelen, smallestAllelelen,sizeratio)
# --- if it was possible to extend it using prodigal --- #
print ("Finished extension at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
#print ORFFoundInMatch
#print strCDS
#print CDSType
isContainedDefinedAllele = False
try:
if ORFFoundInMatch :
alleleStr = strCDS
if match.sbjct_start > match.sbjct_end: #reverse the order if needed
alleleStr = reverseComplement(alleleStr)
lenRatio = float(len(strCDS)) / float(geneLen)
defAllele=[]
if alleleStr in geneDict: #if ORF found is already defined
alleleNumber = geneDict[ alleleStr ]
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
perfectMatchIdAllele.append(alleleNumber)
resultsList.append('EXC2:' + str(alleleNumber) )
else:
#######################
# ADD INFERRED ALLELE # # a new allele that was extended with prodigal
#######################
if(CDSType=='stop codon in match end'):
tagAux = 'INF1:'
elif(CDSType=='start codon in match beggining'):
tagAux = 'INF2:'
elif(CDSType=='bigger than match'):
tagAux = 'INF3:'
elif(CDSType=='same size as match'):
tagAux = 'INF4:'
elif(CDSType=='cds inside match'):
tagAux = 'INF5:'
elif(CDSType=='start codon inside match'):
tagAux = 'INF6:'
else:
tagAux = 'INF7:'
print "infered allele has location : "+(CDSType)
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append( tagAux +"-"+str(alleleI))
print "New allele Infered with prodigal! Adding allele "+ tagAux + str(alleleI) +" to the database\n"
geneDict[alleleStr] = alleleI
resultsList.append( tagAux + str(alleleI) )
orderedAlleleNames.append('allele_' + str(alleleI) + '_' + tagAux[:-1] +"_" +str(os.path.basename(genomeFile)))
# --- add the new allele to the gene fasta --- #
fG = open( geneFile, 'a' )
fG.write('>allele_' + str(alleleI) + '_' + tagAux[:-1] +"_" + str(os.path.basename(genomeFile)) + '\n')
#print alleleStr
fG.write( alleleStr + '\n')
fG.close()
alleleI += 1
# --- remake blast DB --- #
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1,False )
else:
##################
# LNF WTFFF #
##################
geneFile2= os.path.splitext(geneFile)[0] + "LNF2.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
f.write(">Allele\n")
f.write((bmAllele)+"\n")
resultsList.append('LNF2')
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("LNF2")
print "CDS not found"
except:
if ORFFoundInMatch :
alleleStr = strCDS
##################
# LNF WTFFF #
##################
geneFile2= os.path.splitext(geneFile)[0] + "LNF99.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
f.write(">Allele\n")
f.write((bmAllele)+"\n")
resultsList.append('LNF99')
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("LNF99")
print "A problem occurred"
final = (resultsList,perfectMatchIdAllele)
#return (resultsList)
print ("Finished allele calling at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
filepath=os.path.join(basepath , os.path.basename(geneFile)+"_result.txt")
#print filepath
with open(filepath, 'wb') as f:
pickle.dump(final, f)
return True
def callAlleles(argumentList):
geneFile = argumentList[0]
genomesList = argumentList[1]
listOfGenomesDict = argumentList[2]
gene_fp = HTSeq.FastaReader(geneFile)
geneDict = {}
alleleI = 1
orderedAlleleNames=[]
biggestAllelelen=0
smallestAllelelen=99999
for allele in gene_fp:
if allele.seq in geneDict:
print "\nWARNING: this file contains a repeated allele, it should be checked. Ignoring it now!\n", geneFile
else:
if len(allele.seq)>biggestAllelelen:
biggestAllelelen=len(allele.seq)
if len(allele.seq)<smallestAllelelen:
smallestAllelelen=len(allele.seq)
orderedAlleleNames.append(allele.name)
geneDict[ allele.seq ] = alleleI
alleleI += 1
#print geneDict
#print orderedAlleleNames
# --- make 1st blast DB --- #
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 )
geneF = os.path.splitext( geneFile )[0]
blast_out_file = geneF + '.xml'
# list of results - the output of the function
resultsList = []
i = 0
perfectMatchIdAllele=[]
for genomeFile in genomesList:
#print geneDict
currentGenomeDict = listOfGenomesDict[i]
#print genomeFile
#print resultsList
#print geneDict
#print orderedAlleleNames
i+=1 # it has to be incremented here
if genomeFile[-1] == '\n':
genomeFile = genomeFile[:-1]
# ------------------------------ RUNNING BLAST ------------------------------ #
cline = NcbiblastnCommandline(query=genomeFile, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, genomeFile)
# ------ DETERMINING BEST MATCH ------ #
# bestMatch = ['rec.query','hsp', lenRatio]
bestMatch = ['','', 0]
bestMatchContig=''
bestMatchContigLen=''
bestalignlen=0
perfectMatch=False
bmAlleleLen2=0
bmAllele=''
#noAlignment=False
for blast_record in blast_records:
# --- the LNF cases are now called outside de loop --- #
#print blast_record
if perfectMatch==True:
break
try:
#print blast_record.alignments
hspC = blast_record.alignments[0]
if bestMatch[0] == '' and bestMatch[1] == '':
bestMatch[0] = blast_record.query
bestMatch[1] = hspC
except IndexError:
continue
# --- the contig tag is used in the progigal function --- #
contigTag = blast_record.query
# --- brute force parsing of the contig tag - better solution is advisable --- #
j=0
for l in contigTag:
if l == ' ':
break
j+=1
contigTag = contigTag[:j]
contigLen = blast_record.query_letters
#print blast_record.query_id
# --- iterating over all the results to determine the best match --- #
for alignment in blast_record.alignments:
index=orderedAlleleNames.index(alignment.hit_def)
#print alignment.hit_def
for k, v in geneDict.iteritems():
if v == index+1:
bmAlleleLen2= len(k)
if perfectMatch:
break
for match in alignment.hsps:
#print match
scoreRatio = float(match.score) / float(bmAlleleLen2)
#print alignment.hit_def
#print match.identities
#print bmAlleleLen2
#print
#print match.identities
#print len(match.sbjct)
#if #identities is the same as the length of the allele and it has no gaps or N's
if (int(match.identities)==int(bmAlleleLen2) and int(match.identities)==int(len(match.sbjct)) and "N" not in match.query ):
index=orderedAlleleNames.index(alignment.hit_def)
for seq, alleleid in geneDict.iteritems():
if alleleid == index+1:
bmAllele=seq
break
bmAlleleLen= len(bmAllele)
lenratio=float(len(match.query))/float(bmAlleleLen)
bestMatch = [blast_record.query, match, scoreRatio, alignment.hit_def,lenratio,bmAlleleLen]
bestMatchContig=contigTag
perfectMatch=True
index=orderedAlleleNames.index(alignment.hit_def)
bmAlleleLen= len(geneDict.keys()[index])
break
#choose the match with the best score ratio score/length of allele
elif scoreRatio > bestMatch[2]:
index=orderedAlleleNames.index(alignment.hit_def)
#print orderedAlleleNames
#print geneDict
#print orderedAlleleNames
#print alignment.hit_def
#print index
#print geneDict
for seq, alleleid in geneDict.iteritems():
if alleleid == index+1:
bmAllele=seq
break
bmAlleleLen= len(bmAllele)
#print bmAllele
lenratio=float(len(match.query))/float(bmAlleleLen)
bestMatch = [blast_record.query, match, scoreRatio, alignment.hit_def,lenratio,bmAlleleLen]
bestMatchContig=contigTag
bestMatchContigLen=blast_record.query_letters
#print match.query
bestalignlen=alignment.length
if perfectMatch==True:
break
# ---------- ALLELE CALLING AFTER DETERMINING BEST MATCH ---------- #
try:
#print bestMatch[0]
match = bestMatch[1]
#print match
#print match.sbjct
geneLen = bestMatch[5]
alleleStr = match.query
nIdentities = match.identities
idPercent = float(nIdentities) / float(geneLen)
scoreRatio = bestMatch[2]
lenRatio = bestMatch[4]
#print perfectMatch
#print "\nContig best/exact match is :"
#print bestMatchContig +"\n"
except:
#if no best match was found
###################
# LOCUS NOT FOUND #
###################
resultsList.append('LNF:-1') # append result to the list of results
perfectMatchIdAllele.append('LNF')
printinfo(genomeFile,geneFile)
print "Locus not found \n"
continue
if perfectMatch is True:
#if a perfect match was found
if match.sbjct_start > match.sbjct_end: #reverse the order if needed
alleleStr = reverseComplement(alleleStr)
alleleNumber = geneDict[ alleleStr ]
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
if "_" in bestMatch[3]:
a=bestMatch[3].split("_")
perfectMatchIdAllele.append(a[1])
else:
perfectMatchIdAllele.append(bestMatch[3])
resultsList.append('EXC:' + str(alleleNumber) )
elif bestMatch[0] != '' and perfectMatch is not True:
#if a best match was found but it's not an exact match
###########################
# LOCUS ON THE CONTIG TIP #
###########################
if match.query_start ==1 and len(match.query) < geneLen:
resultsList.append('LOT5:-1')
perfectMatchIdAllele.append('LOT5')
printinfo(genomeFile,geneFile)
print "Locus is on the 5' tip of the contig \n"
elif match.query_end == bestMatchContigLen and len(match.query) < bestMatchContigLen:
resultsList.append('LOT3:-1')
perfectMatchIdAllele.append('LOT3')
printinfo(genomeFile,geneFile)
print "Locus is on the 3' tip of the contig \n"
elif bestMatchContigLen <= geneLen:
resultsList.append('LOTSC:-1')
perfectMatchIdAllele.append('LOTSC')
printinfo(genomeFile,geneFile)
#print match.query_start
print "Locus is bigger than the contig \n"
elif 'N' in alleleStr:
#TODO gravar para ficheiro
#####################
# ALLELE NOT FOUND # # N base found!
#####################
geneFile2= os.path.splitext(geneFile)[0] + "LNFN.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+"\n")
f.write((alleleStr) +"\n")
resultsList.append('LNFN:-1')
perfectMatchIdAllele.append('LNFN')
printinfo(genomeFile,geneFile)
print "LNFN, contains N bases! \n"
else:
#removing gaps
alleleStr = alleleStr.replace('-', '')
#lenExtraThresh=int(biggestAllelelen*0.1)
lenExtraThresh=50
#print alleleStr
# --- it might be needed to obtain the reverse complement of the allele string --- #
if match.sbjct_start > match.sbjct_end:
alleleStr = reverseComplement(alleleStr)
#if alleleStr in geneDict: #if best match without gaps is already defined, example: best match allele was already defined but without gaps it's equal to a NA added
# alleleNumber = geneDict[ alleleStr ]
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
# perfectMatchIdAllele.append("EXC2-"+str(alleleNumber))
# resultsList.append('EXC2:' + str(alleleNumber) )
#else: #check if best match without gaps are contained inside an already defined allele
isContainedDefinedAllele = False
#print geneDict.keys()[0]
definedAllele=''
definedAlleleName=''
for k in geneDict.keys():
if alleleStr in k:
definedAllele=k
#print alleleStr
isContainedDefinedAllele = True
definedAlleleName=geneDict.get(k)
break
if isContainedDefinedAllele and int(len(match.query))<=int(len(definedAllele))+lenExtraThresh and int(len(match.query))>=int(len(definedAllele))-lenExtraThresh :
#allele without gaps is contained in a defined allele
#best match with gaps has same size +1/-1 base as the defined allele
#print int(len(definedAllele)), int(len(match.sbjct))
if int(len(alleleStr))==int(len(definedAllele)): # if match without gaps has same size as the defined allele
tagAux = 'NA1:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA1-"+str(alleleI))
elif int(len(alleleStr))==int(len(definedAllele))-1 : # if match without gaps has minus one base than the defined allele
tagAux = 'NA2:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA2-"+str(alleleI))
#elif int(len(alleleStr))==int(len(definedAllele))+1 : # if match without gaps has plus one base than the defined allele
# tagAux = 'NA3:'
# printinfo(genomeFile,geneFile)
# perfectMatchIdAllele.append("NA3-"+str(alleleI))
else: # if match without gaps has more than one base missing comparing to the defined allele
tagAux = 'NA4:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA4-"+str(alleleI))
#TODO catch +1 and others
print "New allele found! Adding allele "+ tagAux + str(alleleI) +" to the database"
geneDict[alleleStr] = alleleI
resultsList.append( tagAux + str(alleleI) )
orderedAlleleNames.append('allele_' + str(alleleI) + '_' + tagAux[:-1] +'_' + str(os.path.basename(genomeFile)))
# --- add the new allele to the gene fasta --- #
fG = open( geneFile, 'a' )
fG.write('>allele_' + str(alleleI) + '_' + tagAux[:-1] +'_' + str(os.path.basename(genomeFile)) + '\n')
fG.write( alleleStr + '\n')
fG.close()
alleleI += 1
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 )
#if best match is not contained in an already defined allele, check if it has similar size with the match allele and has 0.8 similarity
elif not isContainedDefinedAllele and idPercent > 0.8 and int(len(match.query))<=int(geneLen)+lenExtraThresh and int(len(match.query))>=int(geneLen)-lenExtraThresh :
#best match with gaps has 80% identity
#best match with gaps is the same size or +1/-1 as the defined allele
ratio=float(len(alleleStr)) / float(geneLen)
if ratio>=0.8 and ratio<=1.2: # if match without gaps has same size as the best match allele and 80%similarity
tagAux = ''
extraleft=0
extraright=0
tS=0
tE=0
#print int(geneLen), len(match.sbjct)
#print match.sbjct
#print match
handle = open(genomeFile, "rU")
record_dict = SeqIO.to_dict(SeqIO.parse(handle, "fasta"))
handle.close()
record= record_dict[bestMatchContig]
#print match.sbjct
#if(int(len(alleleStr))<int(len(match.query)) and int(len(match.query))<int(geneLen)) and int(geneLen)==int(match.sbjct_start): #if best match allele has missing bases, the tips would be cut
#if len(match.sbjct)<geneLen and "-" not in match.sbjct: #if the allele is not fully used against the match, compensate the tips
if (1<int(match.sbjct_start) and 1<int(match.sbjct_end)):
if match.sbjct_start > match.sbjct_end:
extraleft=match.sbjct_end-1
else:
extraleft=match.sbjct_start-1
if (int(geneLen)>int(match.sbjct_start) and int(geneLen)>int(match.sbjct_end) ): # if 3' tip bases of the allele are missing on the match
if match.sbjct_start > match.sbjct_end:
extraright=geneLen-match.sbjct_start
else:
extraright=geneLen-match.sbjct_end
#print extraleft, extraright
if match.sbjct_start > match.sbjct_end:
tS=match.query_start-extraright-1
tE=match.query_end+extraleft
alleleStr=str(record.seq[tS:tE])
alleleStr = reverseComplement(alleleStr)
else:
tS=match.query_start-extraleft-1
tE=match.query_end+extraright
alleleStr=str(record.seq[tS:tE])
tagAux = 'NA5:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA5-"+str(alleleI))
print "New allele found! Adding allele "+ tagAux + str(alleleI) +" to the database"
geneDict[alleleStr] = alleleI
resultsList.append( tagAux + str(alleleI) )
orderedAlleleNames.append('allele_' + str(alleleI) + '_' + tagAux[:-1] +'_' + str(os.path.basename(genomeFile)))
# --- add the new allele to the gene fasta --- #
fG = open( geneFile, 'a' )
fG.write('>allele_' + str(alleleI) + '_' + tagAux[:-1] +'_' + str(os.path.basename(genomeFile)) + '\n')
fG.write( alleleStr + '\n')
fG.close()
alleleI += 1
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 )
else:
##################
# LNF WTFFF #
##################
geneFile2= os.path.splitext(geneFile)[0] + "LNF3.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
f.write(">Allele\n")
f.write((bmAllele)+"\n")
resultsList.append('LNF3')
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("LNF3")
print "No allele found"
elif isContainedDefinedAllele:
####################
# UNDEFINED ALLELE # # it is contained in another allele
####################
alleleStr=match.query
#if match.sbjct_start > match.sbjct_end: #### - error
#alleleStr = reverseComplement(alleleStr)
resultsList.append('UND:-1')
perfectMatchIdAllele.append("undefined allele")
printinfo(genomeFile,geneFile)
print "Undefined allele \n"
geneFile2= os.path.splitext(geneFile)[0] + "undefined.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
#f.write(">BlastBestMatch"+str(definedAlleleName)+"\n")
#f.write((alleleStr)+"\n")
f.write(">Allele"+str(definedAlleleName)+"\n")
f.write((definedAllele)+"\n")
elif lenRatio < 0.5:
###############
# SMALL MATCH #
###############
resultsList.append('SAC:-1') # don't know what 'SAC' stands for
perfectMatchIdAllele.append('small match')
printinfo(genomeFile,geneFile)
print "lower than 50% match \n"
elif lenRatio < 0.8 and idPercent < 0.5:
#####################
# INCOMPLETE ALLELE # # it was not possible to extend it to at least 80% of the length of the gene
#####################
resultsList.append('INC:-1')
perfectMatchIdAllele.append('allele incomplete')
printinfo(genomeFile,geneFile)
print "Incomplete allele\n"
else:
##################
# LNF WTFFF #
##################
geneFile2= os.path.splitext(geneFile)[0] + "LNF2.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
f.write(">Allele\n")
f.write((bmAllele)+"\n")
resultsList.append('LNF2')
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("LNF2")
print "No allele found"
final = (resultsList,perfectMatchIdAllele)
#return (resultsList)
return final
def main():
try:
input_file = sys.argv[1]
temppath = sys.argv[2]
except IndexError:
print "usage: list_pickle_obj"
argumentList=[]
with open(input_file,'rb') as f:
argumentList = pickle.load(f)
geneFile = argumentList[0]
genomesList = argumentList[1]
basepath=temppath+"/"+os.path.basename(geneFile)
if not os.path.exists(basepath+"/blastdbs/"):
os.makedirs(basepath+"/blastdbs/")
gene_fp = HTSeq.FastaReader(geneFile)
geneDict = {}
alleleI = 1
inverted=False
orderedAlleleNames=[]
biggestAllelelen=0
smallestAllelelen=99999
for allele in gene_fp:
if allele.seq in geneDict:
print "\nWARNING: this file contains a repeated allele, it should be checked. Ignoring it now!\n", geneFile
else:
if len(allele.seq)>biggestAllelelen:
biggestAllelelen=len(allele.seq)
if len(allele.seq)<smallestAllelelen:
smallestAllelelen=len(allele.seq)
orderedAlleleNames.append(str(alleleI))
geneDict[ allele.seq ] = alleleI
alleleI += 1
# --- make 1st blast DB --- #
geneF = os.path.basename(geneFile)
blast_out_file = os.path.dirname(geneFile)+"/blastdbs/"+geneF + '.xml'
# list of results - the output of the function
resultsList = []
i = 0
perfectMatchIdAllele=[]
perfectMatchIdAllele2=[]
genomeDict = {}
genome=-1
print genomesList
for genomeFile in genomesList:
print "_______________________________________________________"
printinfo(genomeFile,geneFile)
g_fp = HTSeq.FastaReader( genomeFile )
for contig in g_fp:
sequence=str(contig.seq)
genomeDict[ contig.name ] = sequence
currentGenomeDict = genomeDict
genome+=1
print ("Blasting alleles on genome at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_out_file = os.path.join(basepath,"blastdbs/"+os.path.basename(geneFile)+ '_List.xml')
Gene_Blast_DB_name = os.path.join(temppath,str(os.path.basename(genomeFile))+"/"+str(os.path.basename(genomeFile))+"_db")
cline = NcbiblastnCommandline(query=geneFile, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, geneFile)
print ("Blasted alleles on genome at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
# ------ DETERMINING BEST MATCH ------ #
bestMatch = ['','', 0]
bestMatchContig=''
bestMatchContigLen=''
bestalignlen=0
perfectMatch=False
bmAlleleLen2=0
bmAllele=''
for blast_record in blast_records:
if perfectMatch==True:
break
try:
hspC = blast_record.alignments[0]
if bestMatch[0] == '' and bestMatch[1] == '':
bestMatch[0] = blast_record.query
bestMatch[1] = hspC
except IndexError:
continue
# --- the contig tag is used in the progigal function --- #
contigTag = blast_record.query
j=0
for l in contigTag:
if l == ' ':
break
j+=1
contigTag = contigTag[:j]
contigLen = blast_record.query_letters
# --- iterating over all the results to determine the best match --- #
for alignment in blast_record.alignments:
contigTag = alignment.hit_def
contigTag=(contigTag.split(" "))[0]
index=orderedAlleleNames.index(str(blast_record.query_id).split("_")[1])
#print alignment.hit_def
for k, v in geneDict.iteritems():
if v == index+1:
bmAlleleLen2= len(k)
if perfectMatch:
break
for match in alignment.hsps:
scoreRatio = float(match.score) / float(bmAlleleLen2)
#if # of identities is the same as the length of the allele and it has no gaps or N's
if (int(match.identities)==int(bmAlleleLen2) and int(match.identities)==int(len(match.query)) and "N" not in match.sbjct and "K" not in match.sbjct and "Y" not in match.sbjct and "R" not in match.sbjct ):
index=orderedAlleleNames.index(str(blast_record.query_id).split("_")[1])
for seq, alleleid in geneDict.iteritems():
if alleleid == index+1:
bmAllele=seq
break
bmAlleleLen= len(bmAllele)
lenratio=float(len(match.sbjct))/float(bmAlleleLen)
bestMatch = [blast_record.query, match, scoreRatio, blast_record.query_id,lenratio,bmAlleleLen]
bestMatchContig=contigTag
perfectMatch=True
index=orderedAlleleNames.index(str(blast_record.query_id).split("_")[1])
bmAlleleLen= len(geneDict.keys()[index])
break
#choose the match with the best score ratio score/length of allele
elif scoreRatio > bestMatch[2]:
index=orderedAlleleNames.index(str(blast_record.query_id).split("_")[1])
for seq, alleleid in geneDict.iteritems():
if alleleid == index+1:
bmAllele=seq
break
bmAlleleLen= len(bmAllele)
lenratio=float(len(match.sbjct))/float(bmAlleleLen)
bestMatch = [blast_record.query, match, scoreRatio, blast_record.query_id,lenratio,bmAlleleLen]
bestMatchContig=contigTag
bestMatchContigLen=len(currentGenomeDict[contigTag])
print contigTag
bestalignlen=alignment.length
if perfectMatch==True:
break
# ---------- ALLELE CALLING AFTER DETERMINING BEST MATCH ---------- #
print ("Finished choosing best match at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
try:
match = bestMatch[1]
bestMatchStart=match.sbjct_start
bestMatchEnd=match.sbjct_end
if match.query_start > match.query_end:
bestMatchEnd=match.sbjct_start
bestMatchStart=match.sbjct_end
geneLen = bestMatch[5]
alleleStr = match.sbjct
nIdentities = match.identities
idPercent = float(nIdentities) / float(geneLen)
scoreRatio = bestMatch[2]
lenRatio = bestMatch[4]
except:
#if no best match was found
###################
# LOCUS NOT FOUND #
###################
resultsList.append('LNF3:-1')
perfectMatchIdAllele.append('LNF')
perfectMatchIdAllele2.append('LNF')
print "Locus not found, no matches \n"
continue
print "is perfect match true?" +str(perfectMatch)
if perfectMatch is True:
#if a perfect match was found (DNA sequence is the same)
try:
alleleNumber = geneDict[ alleleStr ]
except:
alleleStr=reverseComplement(alleleStr)
alleleNumber = geneDict[ alleleStr ]
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
if "_" in bestMatch[3]:
a=bestMatch[3].split("_")
perfectMatchIdAllele.append(a[1])
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(bestMatchStart)+"-"+str(bestMatchEnd)+"&"+"+")
else:
perfectMatchIdAllele.append(bestMatch[3])
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(bestMatchStart)+"-"+str(bestMatchEnd)+"&"+"+")
resultsList.append('EXC:' + str(alleleNumber) )
printinfo(genomeFile,geneFile)
print "Exact match \n"
continue
elif bestMatch[0] != '' and perfectMatch is not True:
#if a best match was found but it's not an exact match
###########################
# LOCUS ON THE CONTIG TIP #
###########################
#check if the match is on the tip of the contig
if bestMatchContigLen <= geneLen:
resultsList.append('LOTSC:-1')
perfectMatchIdAllele.append('LOTSC')
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(bestMatchStart)+"-"+str(bestMatchEnd)+"&"+"+")
printinfo(genomeFile,geneFile)
print "Locus is bigger than the contig \n"
elif match.sbjct_start ==1 and len(match.query) < geneLen:
resultsList.append('LOT5:-1')
perfectMatchIdAllele.append('LOT5')
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(bestMatchStart)+"-"+str(bestMatchEnd)+"&"+"+")
printinfo(genomeFile,geneFile)
print "Locus is on the 5' tip of the contig \n"
elif match.sbjct_end ==1 and len(match.query) < geneLen and match.sbjct_start > match.sbjct_end:
resultsList.append('LOT3:-1')
perfectMatchIdAllele.append('LOT3')
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(bestMatchStart)+"-"+str(bestMatchEnd)+"&"+"+")
printinfo(genomeFile,geneFile)
print "Locus is on the 3' tip of the contig \n"
elif match.sbjct_end == bestMatchContigLen and len(match.query) < bestMatchContigLen:
resultsList.append('LOT3:-1')
perfectMatchIdAllele.append('LOT3')
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(bestMatchStart)+"-"+str(bestMatchEnd)+"&"+"+")
printinfo(genomeFile,geneFile)
print "Locus is on the 3' tip of the contig \n"
elif match.sbjct_start == bestMatchContigLen and len(match.query) < bestMatchContigLen and match.sbjct_start > match.sbjct_end:
resultsList.append('LOT5:-1')
perfectMatchIdAllele.append('LOT5')
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(bestMatchStart)+"-"+str(bestMatchEnd)+"&"+"+")
printinfo(genomeFile,geneFile)
print "Locus is on the 5' tip of the contig \n"
elif 'N' in alleleStr or "K" in alleleStr or "R" in alleleStr or "Y" in alleleStr:
#####################
# ALLELE NOT FOUND #
#####################
# strange base found!
geneFile2= os.path.splitext(geneFile)[0] + "LNFN.fasta"
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+"\n")
f.write((alleleStr) +"\n")
resultsList.append('LNFN:-1')
perfectMatchIdAllele.append('LNFN')
perfectMatchIdAllele2.append('LNFN')
printinfo(genomeFile,geneFile)
print "LNFN, contains strange (N,K,R) bases! \n"
else:
print "new allele?"
alleleStr = alleleStr.replace('-', '')
lenExtraThresh=int(biggestAllelelen*0.2)
#else: #check if best match without gaps are contained inside an already defined allele
isContainedDefinedAllele = False
definedAllele=''
definedAlleleName=''
for k in geneDict.keys():
if alleleStr in k:
definedAllele=k
isContainedDefinedAllele = True
definedAlleleName=geneDict.get(k)
break
print "is contained? " + str(isContainedDefinedAllele)
print idPercent
print geneLen
print lenExtraThresh
print lenRatio
if isContainedDefinedAllele and int(len(match.sbjct))<=int(len(definedAllele))+lenExtraThresh and int(len(match.sbjct))>=int(len(definedAllele))-lenExtraThresh :
#allele without gaps is contained in a defined allele
#best match with gaps has same size +1/-1 base as the defined allele
isnewallele=False
if int(len(alleleStr))==int(len(definedAllele)): # if match without gaps has same size as the defined allele
tagAux = 'NA1:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA1-"+str(alleleI))
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(bestMatchStart)+"-"+str(bestMatchEnd)+"&"+"+")
isnewallele=True
elif int(len(alleleStr))==int(len(definedAllele))-1 : # if match without gaps has minus one base than the defined allele
tagAux = 'NA2:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA2-"+str(alleleI))
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(bestMatchStart)+"-"+str(bestMatchEnd)+"&"+"+")
isnewallele=True
else:
extraleft=0
extraright=0
tS=0
tE=0
handle = open(genomeFile, "rU")
record_dict = SeqIO.to_dict(SeqIO.parse(handle, "fasta"))
handle.close()
record= record_dict[bestMatchContig]
# if match without gaps has more than one base missing comparing to the defined allele
if (1<int(match.query_start) and 1<int(match.query_end)):
if match.query_start > match.query_end:
extraleft=match.query_end-1
else:
extraleft=match.query_start-1
print extraleft, extraright
# if 3' tip bases of the allele are missing on the match
if (int(geneLen)>int(match.query_start) and int(geneLen)>int(match.query_end) ):
if match.query_start > match.query_end:
extraright=geneLen-match.query_start
else:
extraright=geneLen-match.query_end
print extraleft, extraright
if match.sbjct_start > match.sbjct_end:
tE=match.sbjct_start+extraleft
tS=match.sbjct_end-extraright-1
alleleStr=str(record.seq[tS:tE])
alleleStr = reverseComplement(alleleStr)
else:
tS=match.sbjct_start-extraleft-1
tE=match.sbjct_end+extraright
alleleStr=str(record.seq[tS:tE])
print tS
print tE
print "allele is:"
print alleleStr
if tE> bestMatchContigLen:
resultsList.append('LOT3B:-1')
perfectMatchIdAllele.append('LOT3B')
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(tS)+"-"+str(bestMatchContigLen)+"&"+"+")
printinfo(genomeFile,geneFile)
print "Locus is on the 3B' tip of the contig \n"
elif tS<0:
resultsList.append('LOT5B:-1')
perfectMatchIdAllele.append('LOT5B')
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(0)+"-"+str(tE)+"&"+"+")
printinfo(genomeFile,geneFile)
print "Locus is on the 5B' tip of the contig \n"
else:
tagAux = 'NA4:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA4-"+str(alleleI))
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(tS)+"-"+str(tE)+"&"+"+")
isnewallele=True
if isnewallele:
print "New allele found! Adding allele "+ tagAux + str(alleleI) +" to the database"
geneDict[alleleStr] = alleleI
resultsList.append( tagAux + str(alleleI) )
orderedAlleleNames.append(str(alleleI))
# --- add the new allele to the gene fasta --- #
fG = open( geneFile, 'a' )
fG.write('>allele_' + str(alleleI) + '_' + tagAux[:-1] +'_' + str(os.path.basename(genomeFile)) + '\n')
fG.write( alleleStr + '\n')
fG.close()
alleleI += 1
#if best match is not contained in an already defined allele, check if it has similar size with the match allele and has 0.8 similarity
elif not isContainedDefinedAllele and idPercent >= 0.8 and int(len(match.sbjct))<=int(geneLen)+lenExtraThresh and int(len(match.sbjct))>=int(geneLen)-lenExtraThresh :
ratio=float(len(alleleStr)) / float(geneLen)
if ratio>=0.8 and ratio<=1.2: # if match without gaps has same size as the best match allele and 80%similarity
tagAux = ''
extraleft=0
extraright=0
tS=0
tE=0
handle = open(genomeFile, "rU")
record_dict = SeqIO.to_dict(SeqIO.parse(handle, "fasta"))
handle.close()
record= record_dict[bestMatchContig]
#if len(match.sbjct)<geneLen and "-" not in match.sbjct: #if the allele is not fully used against the match, compensate the tips
try:
print match
if (1<int(match.query_start) and 1<int(match.query_end)):
if match.query_start > match.query_end:
extraleft=match.query_end-1
else:
extraleft=match.query_start-1
print extraleft, extraright
if (int(geneLen)>int(match.query_start) and int(geneLen)>int(match.query_end) ): # if 3' tip bases of the allele are missing on the match
if match.query_start > match.query_end:
extraright=geneLen-match.query_start
else:
extraright=geneLen-match.query_end
print extraleft, extraright
if match.sbjct_start > match.sbjct_end:
tE=match.sbjct_start+extraleft
tS=match.sbjct_end-extraright-1
alleleStr=str(record.seq[tS:tE])
alleleStr = reverseComplement(alleleStr)
else:
tS=match.sbjct_start-extraleft-1
tE=match.sbjct_end+extraright
alleleStr=str(record.seq[tS:tE])
print tS
print tE
print "allele is:"
print alleleStr
if tE> bestMatchContigLen:
resultsList.append('LOT3B:-1')
perfectMatchIdAllele.append('LOT3B')
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(tS)+"-"+str(bestMatchContigLen)+"&"+"+")
printinfo(genomeFile,geneFile)
print "Locus is on the 3B' tip of the contig \n"
elif tS<0:
resultsList.append('LOT5B:-1')
perfectMatchIdAllele.append('LOT5B')
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(0)+"-"+str(tE)+"&"+"+")
printinfo(genomeFile,geneFile)
print "Locus is on the 5B' tip of the contig \n"
else:
tagAux = 'NA5:'
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("NA5-"+str(alleleI))
perfectMatchIdAllele2.append(str(bestMatchContig)+"&"+str(tS)+"-"+str(tE)+"&"+"+")
print "New allele found! Adding allele "+ tagAux + str(alleleI) +" to the database"
geneDict[alleleStr] = alleleI
resultsList.append( tagAux + str(alleleI) )
#orderedAlleleNames.append('allele_' + str(alleleI) + '_' + tagAux[:-1] +'_' + str(os.path.basename(genomeFile)))
orderedAlleleNames.append(str(alleleI))
# --- add the new allele to the gene fasta --- #
fG = open( geneFile, 'a' )
fG.write('>allele_' + str(alleleI) + '_' + tagAux[:-1] +'_' + str(os.path.basename(genomeFile)) + '\n')
fG.write( alleleStr + '\n')
fG.close()
alleleI += 1
except Exception as e:
##################
# LNF WTF #
##################
print e
geneFile2= os.path.splitext(geneFile)[0] + "LNF3.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
f.write(">Allele\n")
f.write((bmAllele)+"\n")
resultsList.append('LNF3')
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("LNF3")
perfectMatchIdAllele2.append("LNF3")
print "No allele found"
else:
##################
# LNF WTF2 #
##################
geneFile2= os.path.splitext(geneFile)[0] + "LNF4.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
f.write(">Allele\n")
f.write((bmAllele)+"\n")
resultsList.append('LNF4')
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("LNF4")
perfectMatchIdAllele2.append("LNF4")
print "No allele found"
elif isContainedDefinedAllele:
####################
# UNDEFINED ALLELE # # it is contained in another allele
####################
alleleStr=match.query
resultsList.append('UND:-1')
perfectMatchIdAllele.append("undefined allele")
perfectMatchIdAllele2.append("undefined allele")
printinfo(genomeFile,geneFile)
print "Undefined allele \n"
geneFile2= os.path.splitext(geneFile)[0] + "undefined.fasta"
print geneFile2
"""with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
#f.write(">BlastBestMatch"+str(definedAlleleName)+"\n")
#f.write((alleleStr)+"\n")
f.write(">Allele"+str(definedAlleleName)+"\n")
f.write((definedAllele)+"\n")"""
elif lenRatio < 0.5:
###############
# SMALL MATCH #
###############
resultsList.append('SAC:-1') # don't know what 'SAC' stands for
perfectMatchIdAllele.append('small match')
perfectMatchIdAllele2.append('small match')
printinfo(genomeFile,geneFile)
print "lower than 50% match \n"
elif lenRatio < 0.8 and idPercent < 0.5:
#####################
# INCOMPLETE ALLELE # # it was not possible to extend it to at least 80% of the length of the gene
#####################
resultsList.append('INC:-1')
perfectMatchIdAllele.append('allele incomplete')
perfectMatchIdAllele2.append('allele incomplete')
printinfo(genomeFile,geneFile)
print "Incomplete allele\n"
else:
##################
# LNF WTFFF #
##################
geneFile2= os.path.splitext(geneFile)[0] + "LNF.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">"+ (str(os.path.basename(genomeFile)))+"|"+(str(os.path.basename(geneFile)))+" | "+str(bestMatchContig)+"\n")
f.write((alleleStr) +"\n")
f.write(">Allele\n")
f.write((bmAllele)+"\n")
resultsList.append('LNF5')
printinfo(genomeFile,geneFile)
perfectMatchIdAllele.append("LNF5")
perfectMatchIdAllele2.append("LNF5")
print "Locus not found"
final = (resultsList,perfectMatchIdAllele)
print ("Finished allele calling at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
filepath=os.path.join(temppath , os.path.basename(geneFile)+"_result.txt")
filepath2=os.path.join(temppath , os.path.basename(geneFile)+"_result2.txt")
with open(filepath, 'wb') as f:
pickle.dump(final, f)
with open(filepath2, 'wb') as f:
pickle.dump(perfectMatchIdAllele2, f)
return True
def main():
print ("Starting script at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
try:
input_file = sys.argv[1]
temppath = sys.argv[2]
except IndexError:
print "usage: list_pickle_obj"
argumentList=[]
with open(input_file,'rb') as f:
argumentList = pickle.load(f)
geneFile = argumentList[0]
genomesList = argumentList[1]
#listOfCDSDicts = argumentList[2]
basepath=os.path.join(temppath,os.path.splitext(geneFile)[0])
if not os.path.exists(basepath):
os.makedirs(basepath)
#print geneFile
gene_fp = HTSeq.FastaReader(geneFile)
alleleI = 0
#inverted=False
#orderedAlleleNames=[]
resultsList = []
i = 0
perfectMatchIdAllele=[]
perfectMatchIdAllele2=[]
allelescores=[]
print ("Getting BSR at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
alleleI,allelescores,Gene_Blast_DB_name,alleleList=getBlastScoreRatios(geneFile,basepath)
print ("Finished BSR at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
genome=-1
genomeDict = {}
print ("starting allele call at: "+time.strftime("%H:%M:%S-%d/%m/%Y"))
for genomeFile in genomesList:
print genomeFile
bestmatch=[0,0,False,'',0] #score, score ratio, perfectmatch, key name of the DNA sequence string, allele ID
currentGenomeDict={}
currentCDSDict={}
#currentCDSDict = listOfCDSDicts[i]
filepath=os.path.join(temppath,str(os.path.basename(genomeFile))+"_ORF_Protein.txt")
with open(filepath,'rb') as f:
currentCDSDict = pickle.load(f)
g_fp = HTSeq.FastaReader( genomeFile )
for contig in g_fp:
sequence=str(contig.seq)
genomeDict[ contig.name ] = sequence
currentGenomeDict = genomeDict
#print currentGenomeDict
#alleleI = 0
#alleleProt=''
#for allele in gene_fp: #new db for each allele to blast it against himself
# alleleI+=1
# alleleProt+=">"+str(alleleI)+"\n"+str(translateSeq(allele.seq)+"\n")
#basepath="./blastdbs/temp"+str(os.path.basename(geneFile))
#if not os.path.exists(basepath):
# os.makedirs(basepath)
#with open(basepath+'/protein.fasta', "wb") as f:
# f.write(alleleProt)
#Gene_Blast_DB_name = Create_Blastdb( basepath+'/protein.fasta', 1, True )
genome+=1
listOfCDS=currentCDSDict
genomeProteinfastaPath=os.path.join(temppath,str(os.path.basename(genomeFile)+'_Protein.fasta'))
print ("Blasting alleles on genome at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_out_file = os.path.join(basepath,"blastdbs/"+os.path.basename(geneFile)+ '_List.xml')
#with open(basepath+'/proteinList.fasta', "wb") as f:
# f.write(protList)
#Gene_Blast_DB_name = Create_Blastdb( './temp/proteinList.fasta', 1, True )
cline = NcbiblastpCommandline(query=genomeProteinfastaPath, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
#print cline
#try:
print ("Parse bsr blast at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_records = runBlastParser(cline, blast_out_file, genomeProteinfastaPath)
#except:
# cline = NcbiblastpCommandline(query=genomeProteinfastaPath, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
print ("Blasted alleles on genome at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
for blast_record in blast_records:
for alignment in blast_record.alignments:
for match in alignment.hsps:
#print blast_record.query
#print match
#print alleleI, len(allelescores)
scoreRatio=float(match.score)/float(allelescores[int(alignment.hit_def)-1])
#print scoreRatio
#print alignment.hit_def
cdsStrName=blast_record.query
DNAstr=listOfCDS[">"+cdsStrName]
AlleleDNAstr=alleleList[int(alignment.hit_def)-1]
compare=False
if DNAstr==AlleleDNAstr is False:
try:
DNAstr=reverseComplement(DNAstr)
if DNAstr==AlleleDNAstr is False:
pass
else:
compare=True
except:
pass
else:
compare=True
if "N" in DNAstr or "K" in DNAstr or "R" in DNAstr:
pass
elif(scoreRatio == 1 and bestmatch[2] is False and compare is True):
bestmatch=[match.score,scoreRatio,True,cdsStrName,int(alignment.hit_def),match,len(AlleleDNAstr),blast_record.query_letters]
#print alignment
#print match
elif(scoreRatio == 1 and match.score>bestmatch[0] and compare is True):
bestmatch=[match.score,scoreRatio,True,cdsStrName,int(alignment.hit_def),match,len(AlleleDNAstr),blast_record.query_letters]
#print match
elif(scoreRatio == 1 and bestmatch[2] is False and compare is False):
bestmatch=[match.score,scoreRatio,False,cdsStrName,int(alignment.hit_def),match,len(AlleleDNAstr),blast_record.query_letters]
#print alignment
#print match
elif(scoreRatio == 1 and match.score>bestmatch[0] and compare is False):
bestmatch=[match.score,scoreRatio,False,cdsStrName,int(alignment.hit_def),match,len(AlleleDNAstr),blast_record.query_letters]
#print match
elif(match.score>bestmatch[0] and scoreRatio>0.6 and scoreRatio>bestmatch[1] and bestmatch[2] is False):
#print match.query
#print match.sbjct
#print allelescores
bestmatch=[match.score,scoreRatio,False,cdsStrName,int(alignment.hit_def),match,len(AlleleDNAstr),blast_record.query_letters]
#print match
#print bestmatch
print ("Classifying the match at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
if bestmatch[0]==0 or "N" in AlleleDNAstr or "K" in AlleleDNAstr or "R" in AlleleDNAstr :
#if no best match was found
###################
# LOCUS NOT FOUND #
###################
if bestmatch[0]==0:
resultsList.append('LNF3:-1') # append result to the list of results
perfectMatchIdAllele.append('LNF')
perfectMatchIdAllele2.append('LNF')
#printinfo(genomeFile,geneFile)
print "Locus not found, no matches \n"
else:
resultsList.append('LNFN:-1') # append result to the list of results
perfectMatchIdAllele.append('LNF')
perfectMatchIdAllele2.append('LNF')
#printinfo(genomeFile,geneFile)
print "Locus has strange base (N, K or R) \n"
elif bestmatch[2] is True:
contigname=bestmatch[3]
contigname=contigname.split("&")
matchLocation=contigname[2]
#matchLocation=matchLocation.split("-")
contigname=contigname[0]
alleleStr=listOfCDS[">"+bestmatch[3]]
protSeq,alleleStr,Reversed=translateSeq(alleleStr)
#if a perfect match was found
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
perfectMatchIdAllele.append(str(bestmatch[4]))
if not Reversed:
perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation)+"&"+"+")
else:
perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation)+"&"+"-")
resultsList.append('EXC:' + str(bestmatch[4]) )
else:
match=bestmatch[5]
#print match
geneLen=bestmatch[6]
contigname=bestmatch[3]
#print contigname
contigname=contigname.split("&")
matchLocation=contigname[2]
matchLocation=matchLocation.split("-")
contigname=contigname[0]
seq=currentGenomeDict[ contigname ]
bestMatchContigLen=len(seq)
alleleStr=listOfCDS[">"+bestmatch[3]]
protSeq,alleleStr,Reversed=translateSeq(alleleStr)
print match
print matchLocation
print bestMatchContigLen
# get extra space to the right and left between the contig and match
rightmatchContig=bestMatchContigLen-int(matchLocation[1])
leftmatchContig=int(matchLocation[0])
if Reversed:
aux=rightmatchContig
rightmatchContig=leftmatchContig
leftmatchContig=aux
"""else:
rightmatchContig=bestMatchContigLen-int(matchLocation[0])
leftmatchContig=int(matchLocation[1])"""
print rightmatchContig,leftmatchContig
# get extra space to the right and left between the allele and match
rightmatchAllele=geneLen-(int(match.sbjct_end)*3)
leftmatchAllele=(int(match.sbjct_start)*3)
"""if Reversed:
aux=rightmatchAllele
rightmatchAllele=leftmatchAllele
leftmatchAllele=aux"""
"""else:
rightmatchAllele=geneLen-(int(match.sbjct_start )*3)
leftmatchAllele=(int(match.sbjct_end)*3)"""
#if a best match was found but it's not an exact match
###########################
# LOCUS ON THE CONTIG TIP #
###########################
print rightmatchAllele, leftmatchAllele
print geneLen
#if bestMatchContigLen <= geneLen:
if leftmatchContig<leftmatchAllele and rightmatchContig < rightmatchAllele:
resultsList.append('LOTSC:-1')
perfectMatchIdAllele.append('LOTSC')
perfectMatchIdAllele2.append('LOTSC')
#printinfo(genomeFile,geneFile)
#print match.query_start
print "Locus is bigger than the contig \n"
#if match.query_start ==1 and len(match.query) < geneLen:
elif leftmatchContig<leftmatchAllele:
resultsList.append('LOT3:-1')
perfectMatchIdAllele.append('LOT3')
perfectMatchIdAllele2.append('LOT3')
print "Locus is on the 3' tip of the contig \n"
#elif match.query_end == bestMatchContigLen and len(match.query) < geneLen:
elif rightmatchContig < rightmatchAllele:
resultsList.append('LOT5:-1')
perfectMatchIdAllele.append('LOT5')
perfectMatchIdAllele2.append('LOT5')
print "Locus is on the 5' tip of the contig \n"
else:
#######################
# ADD INFERRED ALLELE # # a new allele
#######################
#print "infered allele has location : "+(CDSType)
#printinfo(genomeFile,geneFile)
tagAux='INF'
perfectMatchIdAllele.append( tagAux +"-"+str(alleleI+1))
#perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1]))
if not Reversed:
perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"+")
else:
perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"-")
print "New allele! Adding allele "+ tagAux + str(alleleI+1) +" to the database\n"
resultsList.append( tagAux + str(alleleI+1) )
#orderedAlleleNames.append('allele_' + str(alleleI) + '_' + tagAux[:-1] +"_" +str(os.path.basename(genomeFile)))
# --- add the new allele to the gene fasta --- #
appendAllele='>allele_' + str(alleleI+1) + '_' + tagAux[:-1] +"_" + str(os.path.basename(genomesList[genome])) + '\n'
fG = open( geneFile, 'a' )
#fG.write('>allele_' + str(alleleI+1) + '_' + tagAux[:-1] +"_" + str(os.path.basename(genomesList[genome])) + '\n')
fG.write(appendAllele)
#print alleleStr
#print listOfCDS
#alleleStr=listOfCDS[">"+bestmatch[3]]
#match=bestmatch[5]
#reverse the order if needed
#if match.sbjct_start > match.sbjct_end:
# alleleStr = reverseComplement(alleleStr)
fG.write( alleleStr + '\n')
fG.close()
fG = open( os.path.join(basepath,str(os.path.basename(geneFile)+'_protein2.fasta')), 'w' )
#fG.write('>allele_' + str(alleleI+1) + '_' + tagAux[:-1] +"_" + str(os.path.basename(genomesList[genome])) + '\n')
fG.write('>'+str(alleleI+1)+'\n'+str(protSeq) + '\n')
fG.close()
fG = open( os.path.join(basepath,str(os.path.basename(geneFile)+'_protein.fasta')), 'a' )
#fG.write('>allele_' + str(alleleI+1) + '_' + tagAux[:-1] +"_" + str(os.path.basename(genomesList[genome])) + '\n')
fG.write('>'+str(alleleI+1)+'\n'+str(protSeq) + '\n')
#print alleleStr
fG.close()
#print listOfCDS
#alleleStr=listOfCDS[">"+bestmatch[3]]
match=bestmatch[5]
#alleleI += 1
# --- remake blast DB --- #
alleleList.append(alleleStr)
Gene_Blast_DB_name = Create_Blastdb( os.path.join(basepath,str(os.path.basename(geneFile)+'_protein.fasta')), 1, True )
print os.path.join(basepath,str(os.path.basename(geneFile)+'_protein.fasta'))
genefile2= os.path.join(basepath,str(os.path.basename(geneFile)+'_protein2.fasta'))
Gene_Blast_DB_name2 = Create_Blastdb( genefile2, 1, True )
print ("Re-calculating BSR at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
alleleI,allelescores,alleleList=reDogetBlastScoreRatios(genefile2,basepath,alleleI,allelescores,Gene_Blast_DB_name2,alleleList)
print allelescores
print ("Done Re-calculating BSR at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
#x=y
#shutil.rmtree(basepath)
final = (resultsList,perfectMatchIdAllele)
#return (resultsList)
print ("Finished allele calling at : "+time.strftime("%H:%M:%S-%d/%m/%Y"))
filepath=os.path.join(temppath , os.path.basename(geneFile)+"_result.txt")
filepath2=os.path.join(temppath , os.path.basename(geneFile)+"_result2.txt")
#print filepath
with open(filepath, 'wb') as f:
pickle.dump(final, f)
with open(filepath2, 'wb') as f:
pickle.dump(perfectMatchIdAllele2, f)
shutil.rmtree(basepath)
return True
def getBlastScoreRatios(genefile, basepath, doAll, verbose, blastPath):
if verbose:
def verboseprint(*args):
for arg in args:
print(arg),
print
else:
verboseprint = lambda *a: None # do-nothing function
#gene_fp = HTSeq.FastaReader(genefile)
allelescores = []
alleleProt = ''
alleleAllProt = ''
alleleList = []
alleleI = 0
alleleIlist = []
listAllelesNames = []
# calculate bsr for each allele
for allele in SeqIO.parse(genefile, "fasta", generic_dna):
# usually first allele name is just >1 and after that it has >gene_id_genome
aux = allele.id.split("_")
if len(aux) < 2:
alleleI = int(aux[0])
else:
alleleI = int(aux[-1])
# try to translate the allele
alleleIlist.append(alleleI)
alleleList.append(str(allele.seq))
listAllelesNames.append(allele.id)
translatedSequence, x, y = translateSeq(allele.seq)
if translatedSequence == '':
print("cannot translate allele on bsr calculation")
pass
# calculate BSR for the allele
else:
alleleProt = ">" + str(alleleI) + "\n" + str(translatedSequence +
"\n")
alleleAllProt += ">" + str(alleleI) + "\n" + str(
translatedSequence + "\n")
proteinfastaPath = os.path.join(
basepath, str(os.path.basename(genefile) + '_protein2.fasta'))
# new db for each allele to blast it against himself
with open(proteinfastaPath, "w") as f:
f.write(alleleProt)
Gene_Blast_DB_name = Create_Blastdb(proteinfastaPath, 1, True)
# if bsr hasn't been calculated, do the BLAST
if doAll:
blast_out_file = os.path.join(basepath, 'blastdbs/temp.xml')
verboseprint("Starting Blast alleles at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
# --- get BLAST score ratio --- #
cline = NcbiblastpCommandline(cmd=blastPath,
query=proteinfastaPath,
db=Gene_Blast_DB_name,
evalue=0.001,
out=blast_out_file,
outfmt=5,
num_threads=1)
allelescore = 0
blast_records = runBlastParser(cline, blast_out_file)
verboseprint("Blasted alleles at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
for blast_record in blast_records:
for alignment in blast_record.alignments:
for match in alignment.hsps:
allelescores.append(int(match.score))
geneScorePickle = os.path.abspath(genefile) + '_bsr.txt'
verboseprint("________")
# ~ var=[alleleI,allelescores]
var = dict(zip(alleleIlist, allelescores))
with open(geneScorePickle, 'wb') as f:
pickle.dump(var, f)
# bsr had already been calculated, load it to memory
else:
geneScorePickle = os.path.abspath(genefile) + '_bsr.txt'
with open(geneScorePickle, 'rb') as f:
var = pickle.load(f)
# ~ allelescores=var[1]
proteinfastaPath = os.path.join(
basepath, str(os.path.basename(genefile) + '_protein.fasta'))
with open(proteinfastaPath, "w") as f:
f.write(alleleAllProt)
# returning all allele BSR scores and list of alleles for this gene
return var, alleleList, listAllelesNames
def main():
parser = argparse.ArgumentParser(
description=
"Given an ffn file, recovers the genes that are not paralogs and have a size bigger than the g parameter provided"
)
parser.add_argument('-i',
nargs='?',
type=str,
help='ffn file',
required=True)
parser.add_argument('-l',
nargs='?',
type=int,
help='int minimum length',
required=True)
parser.add_argument(
'--cpu',
nargs='?',
type=int,
help="Number of cpus, if over the maximum uses maximum -2",
required=False)
parser.add_argument('-p',
nargs='?',
type=str,
help="file with protein",
required=False,
default=False)
parser.add_argument('-o',
nargs='?',
type=str,
help="output filename",
required=False,
default=False)
parser.add_argument('-b',
nargs='?',
type=str,
help="BLAST full path",
required=False,
default='blastp')
parser.add_argument('--bsr',
nargs='?',
type=float,
help="minimum BSR similarity",
required=False,
default=0.6)
parser.add_argument("-v",
"--verbose",
help="increase output verbosity",
dest='verbose',
action="store_true",
default=False)
args = parser.parse_args()
genes = args.i
sizethresh = args.l
cpuToUse = args.cpu
proteinFIlePath = args.p
outputFIlePath = args.o
BlastpPath = args.b
bsr = args.bsr
verbose = args.verbose
if verbose:
def verboseprint(*args):
for arg in args:
print(arg),
print
else:
verboseprint = lambda *a: None
starttime = "\nStarting Script at : " + time.strftime("%H:%M:%S-%d/%m/%Y")
verboseprint("\nStarting Script at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
verboseprint("Checking Blast installed... " + str(which(BlastpPath)))
# translate to protein and create new file
abspath = os.path.abspath(genes)
filename = os.path.basename(genes)
abspath = abspath.replace(filename, '')
proteinfile = os.path.join(abspath, 'proteins.fasta')
geneDict = {}
protDict = {}
orderedprotDict = collections.OrderedDict()
alreadyIn = []
totalgenes = 0
repeatedgenes = 0
smallgenes = 0
nottranslatable = 0
verboseprint("Checking translatability of the loci:\n")
if not proteinFIlePath:
# print "not passing steps"
with open(proteinfile, "w") as f:
#g_fp = HTSeq.FastaReader(genes)
for gene in SeqIO.parse(genes, "fasta", generic_dna):
dnaseq = str(gene.seq)
protseq, seq, y = translateSeq(dnaseq, gene.id)
totalgenes += 1
if len(protseq) > 1:
if str(protseq) in alreadyIn:
repeatedgenes += 1
elif len(str(seq)) < sizethresh:
smallgenes += 1
else:
alreadyIn.append(str(protseq))
protname = ">" + str(gene.id) + "\n"
f.write(protname + str(protseq) + "\n")
protDict[protname] = str(protseq)
geneDict[str(gene.name)] = dnaseq
else:
nottranslatable += 1
continue
verboseprint(
str(nottranslatable) + " not translatable out of " +
str(totalgenes))
verboseprint("\nChecking if repeated protein sequences:\n")
orderedprotList = []
orderedprotList = sorted(protDict.items(),
key=lambda x: len(x[1]),
reverse=True)
i = 0
while i < len(orderedprotList):
elem = orderedprotList[i]
orderedprotDict[elem[0]] = elem[1]
i += 1
verboseprint(
str(repeatedgenes) + " repeated loci out of " + str(totalgenes))
verboseprint(
str(smallgenes) + " loci out of " + str(totalgenes) +
" smaller than " + str(sizethresh) + "bp")
verboseprint("\nprotein file created\n")
# first step - remove genes contained in other genes or 100% equal genes
# list of results - the output of the function
resultsList = []
auxDict = {}
#g_fp = HTSeq.FastaReader(proteinfile)
g = 0
j = 0
verboseprint(
"Checking if protein sequences are contained in others...")
# for each gene from all the annotated genes - starting with an empty dictionary, only add a new gene if the "to be added gene" is not contained or equal to a gene already added to the dictionary
auxprot = []
for elem in orderedprotDict.items():
contained = False
prot = str(elem[1])
if any(prot in x for x in auxprot):
g += 1
contained = True
else:
auxDict[elem[1]] = elem[0]
auxprot.append(str(elem[1]))
j += 1
verboseprint(str(g) + " loci are contained in other genes\n")
# overwrite the original file, obtaining a new file with unique genes
with open(proteinfile, "w") as f:
allsequences = ''
for k, v in auxDict.items():
allsequences += v + k + "\n"
f.write(allsequences)
else:
# print "passed steps"
proteinfile = proteinFIlePath
totalgenes = 0
smallgenes = 0
#g_fp = HTSeq.FastaReader(genes)
proteinfile = proteinFIlePath
for gene in SeqIO.parse(genes, "fasta", generic_dna):
#for gene in g_fp:
dnaseq = str(gene.seq)
protname = ">" + str(gene.id) + "\n"
# protDict[protname] = str(protseq)
geneDict[str(gene.name)] = dnaseq
verboseprint("Starting Blast")
# print "Blasting the total of "+ str(len(auxDict.keys())) + " loci"
geneFile = os.path.abspath(proteinfile)
Gene_Blast_DB_name = Create_Blastdb(geneFile, 1, True)
geneF = os.path.splitext(geneFile)[0]
blast_out_file = geneF + '.xml'
# ------------------------------ RUNNING BLAST ------------------------------ #
if cpuToUse:
cline = NcbiblastpCommandline(cmd=BlastpPath,
query=geneFile,
db=Gene_Blast_DB_name,
evalue=0.001,
out=blast_out_file,
outfmt=5,
num_threads=int(cpuToUse))
else:
cline = NcbiblastpCommandline(cmd=BlastpPath,
query=geneFile,
db=Gene_Blast_DB_name,
evalue=0.001,
out=blast_out_file,
outfmt=5,
num_threads=1)
blast_records = runBlastParser(cline, blast_out_file)
verboseprint("Finished blast")
toRemove = []
genesToKeep = []
log = ["removed\tcause\texplanation"]
for blast_record in blast_records:
allelename = blast_record.query
allelename = allelename.split(" ")
allelename = allelename[0]
alleleLength = len(geneDict[allelename])
try:
# if gene A is not on the toRemove list yet, add to genesToKeep list
if str(blast_record.query) not in toRemove:
genesToKeep.append(blast_record.query)
i = 0
# if first alignement is not against self, gene B is bigger than gene A and very simillar - remove gene A from genesToKeep and add gene B instead
if not str(blast_record.query) == str(
(blast_record.alignments[0]).hit_def):
genesToKeep.remove(str(blast_record.query))
toRemove.append(str(blast_record.query))
log.append(
str(blast_record.query) + "\t" +
str((blast_record.alignments[0]).hit_def) + "\t" +
"2 is first best match")
# if gene B is not on the toRemove list, add to genesToKeep list
if str(
(blast_record.alignments[0]).hit_def) not in toRemove:
genesToKeep.append(
str((blast_record.alignments[0]).hit_def))
raise
selfblastscore = (((blast_record.alignments[0]).hsps)[0]).score
while i < len(blast_record.alignments):
align = blast_record.alignments[i]
match = (align.hsps)[0]
scoreRatio = float(match.score) / float(selfblastscore)
alleleLength2 = len(geneDict[str(align.hit_def)])
# if good match and gene B not in toremove list
if (scoreRatio > bsr and
not str(align.hit_def) == str(blast_record.query)
and str(align.hit_def) not in toRemove):
# if gene B is bigger than gene A, keep bigger gene B
if alleleLength2 > alleleLength:
genesToKeep.append(str(align.hit_def))
genesToKeep.remove(str(blast_record.query))
toRemove.append(str(blast_record.query))
log.append(
str(blast_record.query) + "\t" +
str(align.hit_def) + "\t" +
"2 is bigger and bsr >" + str(bsr))
raise
# else add gene B to toremove list
elif str(align.hit_def) in genesToKeep:
genesToKeep.remove(str(align.hit_def))
toRemove.append(str(align.hit_def))
log.append(
str(align.hit_def) + "\t" +
str(blast_record.query) + "\t" +
"2 is bigger and bsr >" + str(bsr))
i += 1
# else gene A is on toRemove list, add all similar genes (not in genesToKeep) list to the toRemove list
else:
i = 0
selfblastscore = 0
for align in blast_record.alignments:
if not (str(align.hit_def) == str(blast_record.query)):
selfblastscore = ((align.hsps)[0]).score
# print "gene "+str(align.hit_def)+" is larger than gene "+str(blast_record.query)
raise
while i < len(blast_record.alignments):
align = blast_record.alignments[i]
match = (align.hsps)[0]
scoreRatio = float(match.score) / float(selfblastscore)
if align.hit_def not in genesToKeep and not str(
align.hit_def) == str(
blast_record.query) and scoreRatio > bsr:
toRemove.append(align.hit_def)
log.append(
str(align.hit_def) + "\t" +
str(blast_record.query) + "\t" +
"2 was on the removed list and bsr >" + str(bsr))
else:
pass
i += 1
except Exception as e:
# print e
pass
genesToKeep = list(set(genesToKeep))
toRemove = list(set(toRemove))
s = set(toRemove)
notcommonToKeep = [x for x in genesToKeep if x not in s]
pathfiles = os.path.dirname(geneFile)
pathfiles = pathfiles + "/"
listfiles = []
#g_fp = HTSeq.FastaReader(genes)
removedparalogs = 0
removedsize = 0
totalgenes = 0
rest = 0
concatenatedFile = ''
schema_folder_path = os.path.join(pathfiles, 'schema_seed')
if not os.path.exists(
schema_folder_path) and not proteinFIlePath and not outputFIlePath:
os.makedirs(schema_folder_path)
elif not proteinFIlePath and outputFIlePath:
os.makedirs(outputFIlePath)
for contig in SeqIO.parse(genes, "fasta", generic_dna):
totalgenes += 1
#name = contig.name + " " + contig.descr
name2 = contig.id
# print name2
if name2 not in toRemove and name2 in genesToKeep:
if int(len(contig.seq)) > sizethresh:
namefile = contig.name
namefile = namefile.replace("|", "_")
namefile = namefile.replace("_", "-")
namefile = namefile.replace("(", "")
namefile = namefile.replace(")", "")
namefile = namefile.replace("'", "")
namefile = namefile.replace("\"", "")
namefile = namefile.replace(":", "")
if not proteinFIlePath and not outputFIlePath:
newFile = os.path.join(schema_folder_path,
namefile + ".fasta")
listfiles.append(newFile)
with open(newFile, "w") as f:
f.write(">" + namefile + "_1\n" + str(contig.seq) +
"\n")
elif not proteinFIlePath and outputFIlePath:
newFile = os.path.join(outputFIlePath, namefile + ".fasta")
listfiles.append(newFile)
with open(newFile, "w") as f:
f.write(">" + namefile + "_1\n" + str(contig.seq) +
"\n")
else:
concatenatedFile += ">" + contig.id + " \n" + str(
contig.seq) + "\n"
rest += 1
else:
removedsize += 1
else:
removedparalogs += 1
if proteinFIlePath and outputFIlePath:
with open(outputFIlePath, "w") as f:
f.write(concatenatedFile)
elif not proteinFIlePath and outputFIlePath:
get_Short(listfiles)
verboseprint("\nRemoved " + str(removedparalogs) +
" with a high similarity (BSR>" + str(bsr) + ")")
print("Total of " + str(rest) + " loci that constitute the schema")
os.remove(proteinfile)
# create short folder
else:
# ~ with open("schemacreation.log", "wb") as f:
# ~ for elem in log:
# ~ f.write(str(elem)+"\n")
get_Short(listfiles)
verboseprint("\nRemoved " + str(removedparalogs) +
" with a high similarity (BSR>" + str(bsr) + ")")
print("Total of " + str(rest) + " loci that constitute the schema")
os.remove(proteinfile)
shutil.rmtree(os.path.join(pathfiles, 'blastdbs'))
os.remove(blast_out_file)
verboseprint(starttime)
verboseprint("Finished Script at : " + time.strftime("%H:%M:%S-%d/%m/%Y"))
def main():
try:
input_file = sys.argv[1]
temppath = sys.argv[2]
blastPath = sys.argv[3]
verbose = sys.argv[4]
bsrTresh = sys.argv[5]
if verbose == 'True':
verbose = True
else:
verbose = False
except IndexError:
print(
"Error starting the callAlleleles_protein3 script. usage: list_pickle_obj"
)
bsrTresh = float(bsrTresh)
argumentList = []
with open(input_file, 'rb') as f:
argumentList = pickle.load(f)
if verbose:
def verboseprint(*args):
for arg in args:
print(arg),
print
else:
verboseprint = lambda *a: None
geneFile = argumentList[0]
verboseprint("Using gene: " + str(geneFile))
shortgeneFile = os.path.join(os.path.dirname(argumentList[0]), "short",
os.path.basename(argumentList[0]))
shortgeneFile = shortgeneFile.replace(".fasta", "_short.fasta")
genomesList = argumentList[1]
genesList = argumentList[2]
newListgenes = []
with open(genesList, 'r') as gene_fp:
for gene in gene_fp:
gene = gene.rstrip('\n')
gene = gene.rstrip('\r')
newListgenes.append(gene)
statusbar = float(newListgenes.index(str(geneFile))) / len(newListgenes)
locusnumber = (newListgenes.index(str(geneFile)))
totalocusnumber = len(newListgenes)
basepath = os.path.join(temppath, os.path.splitext(geneFile)[0])
print("\rProcessing " + os.path.basename(geneFile) + ". Start " +
time.strftime("%H:%M:%S-%d/%m/%Y") + " Locus " + str(locusnumber) +
" of " + str(totalocusnumber) + ". Done " +
str(int(statusbar * 100)) + "%.",
end="")
if not os.path.exists(basepath):
os.makedirs(basepath)
#gene_fp = HTSeq.FastaReader(geneFile)
fullAlleleList = []
fullAlleleNameList = []
alleleI = 0
# get full list of alleles from main gene file and last allele number id
for allele in SeqIO.parse(geneFile, "fasta", generic_dna):
aux = allele.id.split("_")
if len(aux) < 2:
alleleI = int(aux[0])
else:
alleleI = int(aux[-1])
fullAlleleList.append(str(allele.seq))
fullAlleleNameList.append(allele.id)
resultsList = []
i = 0
perfectMatchIdAllele = []
perfectMatchIdAllele2 = []
allelescores = []
listShortAllelesNames = []
verboseprint("Getting BSR at : " + time.strftime("%H:%M:%S-%d/%m/%Y"))
geneScorePickle = os.path.abspath(shortgeneFile) + '_bsr.txt'
# check if bsr as arealdy been calculated and recalculate it if necessary
if os.path.isfile(geneScorePickle):
allelescores, alleleList, listShortAllelesNames = getBlastScoreRatios(
shortgeneFile, basepath, False, verbose, blastPath)
else:
allelescores, alleleList, listShortAllelesNames = getBlastScoreRatios(
shortgeneFile, basepath, True, verbose, blastPath)
verboseprint("Finished BSR at : " + time.strftime("%H:%M:%S-%d/%m/%Y"))
verboseprint("starting allele call blast at: " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
for genomeFile in genomesList:
verboseprint(genomeFile)
bestmatch = [
0, 0, False, '', 0
] # score, score ratio, perfectmatch, key name of the DNA sequence string, allele ID
currentGenomeDict = {}
currentCDSDict = {}
# load the CDS from the genome to a dictionary
filepath = os.path.join(
temppath,
str(os.path.basename(genomeFile)) + "_ORF_Protein.txt")
with open(filepath, 'rb') as f:
currentCDSDict = pickle.load(f)
try:
intersection = set(fullAlleleList).intersection(
currentCDSDict.values())
intersection = list(intersection)
if len(intersection) > 1:
perfectMatchIdAllele.append('NIPHEM')
perfectMatchIdAllele2.append('NIPHEM')
verboseprint(
os.path.basename(genomeFile) + " has " +
str(len(intersection)) + " multiple exact match : " +
os.path.basename(geneFile) +
" MULTIPLE ALLELES as EXACT MATCH")
raise ValueError("MULTIPLE ALLELES as EXACT MATCH")
elif len(intersection) == 1:
alleleStr = intersection[0]
# it doenst return both keys with equal values
# ~ elem=currentCDSDict.keys()[currentCDSDict.values().index(alleleStr)]
elem = [
key for key, value in currentCDSDict.items()
if value == alleleStr
]
if len(elem) > 1:
perfectMatchIdAllele.append('NIPHEM')
perfectMatchIdAllele2.append('NIPHEM')
verboseprint(
os.path.basename(genomeFile) + " has " +
str(len(intersection)) + " multiple exact match : " +
os.path.basename(geneFile) +
" MULTIPLE ALLELES as EXACT MATCH")
raise ValueError("MULTIPLE ALLELES as EXACT MATCH")
contigname = elem[0].split("&")
matchLocation = contigname[2]
# starting CDS base need to be +1
matchLocation = matchLocation.split("-")
matchLocation = [
int(matchLocation[0]) + 1,
int(matchLocation[1])
]
contigname = (contigname[0]).replace(">", "")
alleleName = ''
alleleMatchid = 0
alleleName = fullAlleleNameList[fullAlleleList.index(
alleleStr)]
alleleMatchid = int((alleleName.split("_"))[-1])
perfectMatchIdAllele.append(str(alleleMatchid))
if matchLocation[0] > matchLocation[1]:
perfectMatchIdAllele2.append(
str(contigname) + "&" + str(matchLocation[0]) + "-" +
str(matchLocation[1]) + "&" + "-")
else:
perfectMatchIdAllele2.append(
str(contigname) + "&" + str(matchLocation[0]) + "-" +
str(matchLocation[1]) + "&" + "+")
# check if atributed allele is contained or contains
try:
containedInfo = (alleleName.split("_"))[1]
except:
containedInfo = ''
if containedInfo == "CD":
resultsList.append([(os.path.basename(genomeFile)),
str(alleleMatchid),
containedInfo.rstrip()])
elif containedInfo == "CS":
resultsList.append([(os.path.basename(genomeFile)),
str(alleleMatchid),
containedInfo.rstrip()])
else:
pass
raise ValueError("EQUAL")
except Exception as e:
# ~ exc_type, exc_obj, exc_tb = sys.exc_info()
# ~ fname = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]
# ~ print(exc_tb.tb_lineno)
# ~ print e
continue
else:
verboseprint("Blasting alleles on genome at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_out_file = os.path.join(
basepath,
"blastdbs/" + os.path.basename(geneFile) + '_List.xml')
Gene_Blast_DB_name = os.path.join(
temppath,
str(os.path.basename(genomeFile)) + "/" +
str(os.path.basename(genomeFile)) + "_db")
proteinfastaPath = os.path.join(
basepath,
str(os.path.basename(shortgeneFile) + '_protein.fasta'))
# blast the genome CDS against the translated locus
# cline = NcbiblastpCommandline(query=proteinfastaPath, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5,max_target_seqs=10,max_hsps_per_subject=10)
# 2.2.28 up
cline = NcbiblastpCommandline(cmd=blastPath,
query=proteinfastaPath,
db=Gene_Blast_DB_name,
evalue=0.001,
out=blast_out_file,
outfmt=5,
max_target_seqs=10,
max_hsps=10,
num_threads=1)
blast_records = runBlastParser(cline, blast_out_file)
verboseprint("Blasted alleles on genome at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
alleleSizes = []
for allele in fullAlleleList:
alleleSizes.append(len(allele))
biggestSizeAllele = max(alleleSizes)
# get mode allele size
moda = max(set(alleleSizes), key=alleleSizes.count)
contador = Counter(alleleSizes).most_common()
# if most common allele size appears 1 time, get first allele size
if (contador[0])[1] == 1:
moda = alleleSizes[0]
try:
# iterate through the blast results
for blast_record in blast_records:
locationcontigs = []
for alignment in blast_record.alignments:
# select the best match
for match in alignment.hsps:
# query id comes with query_id, not name of the allele
alleleMatchid = int(
(blast_record.query_id.split("_"))[-1])
# ~ scoreRatio=float(match.score)/float(allelescores[int(alleleMatchid)-1])
# query_id starts with 1
alleleMatchid2 = ((
listShortAllelesNames[alleleMatchid -
1]).split("_"))[-1]
scoreRatio = float(match.score) / float(
allelescores[int(alleleMatchid2)])
cdsStrName = (alignment.title.split(" "))[1]
#DNAstr = str(currentCDSDict[">" + cdsStrName])
AlleleDNAstr = alleleList[int(alleleMatchid) - 1]
if scoreRatio >= bsrTresh:
locationcontigs.append(cdsStrName)
# select the best match from BLAST results
if scoreRatio == 1 and match.score > bestmatch[0]:
bestmatch = [
match.score, scoreRatio, False, cdsStrName,
int(alleleMatchid), match,
len(AlleleDNAstr)
]
elif (match.score > bestmatch[0]
and scoreRatio >= bsrTresh
and scoreRatio > bestmatch[1]
and bestmatch[2] is False):
bestmatch = [
match.score, scoreRatio, False, cdsStrName,
int(alleleMatchid), match,
len(AlleleDNAstr)
]
verboseprint("Classifying the match at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
# if no best match was found it's a Locus Not Found
# check for ambiguious bases
if not bestmatch[0] == 0:
alleleStr = currentCDSDict[">" + bestmatch[3]]
listFoundAmbiguities = []
listambiguousBases = [
'K', 'M', 'R', 'Y', 'S', 'W', 'B', 'V', 'H', 'D', 'X',
'N', '-', '.'
]
listFoundAmbiguities = [
e for e in listambiguousBases if e in alleleStr
]
if bestmatch[0] == 0 or len(listFoundAmbiguities) > 0:
###################
# LOCUS NOT FOUND #
###################
if bestmatch[0] == 0:
perfectMatchIdAllele.append('LNF')
perfectMatchIdAllele2.append('LNF')
verboseprint("Locus not found, no matches \n")
else:
perfectMatchIdAllele.append('LNF')
perfectMatchIdAllele2.append('LNF')
verboseprint("Locus has strange base \n")
# if more than one BSR >0.6 in two different CDSs it's a Non Paralog Locus
elif len(list(set(locationcontigs))) > 1:
verboseprint("NIPH", "")
perfectMatchIdAllele.append('NIPH')
perfectMatchIdAllele2.append('NIPH')
for elem in locationcontigs:
verboseprint(elem)
# if match with BSR >0.6 and not equal DNA sequences
else:
# load the contig info of the genome to a dictionary
#g_fp = HTSeq.FastaReader(genomeFile)
for contig in SeqIO.parse(genomeFile, "fasta",
generic_dna):
currentGenomeDict[contig.id] = len(str(contig.seq))
match = bestmatch[5]
geneLen = bestmatch[6]
alleleStr = currentCDSDict[">" + bestmatch[3]]
contigname = bestmatch[3]
contigname = contigname.split("&")
matchLocation = contigname[2]
matchLocation = matchLocation.split("-")
matchLocation = [
int(matchLocation[0]) + 1, matchLocation[1]
]
contigname = contigname[0]
bestMatchContigLen = currentGenomeDict[contigname]
protSeq, alleleStr, Reversed = translateSeq(alleleStr)
# get extra space to the right and left between the allele and match and check if it's still inside the contig
rightmatchAllele = geneLen - (
(int(match.query_end) + 1) * 3)
leftmatchAllele = ((int(match.query_start) - 1) * 3)
# ~ if Reversed swap left and right contig extra
if int(matchLocation[1]) < int(matchLocation[0]):
rightmatchContig = bestMatchContigLen - int(
matchLocation[0])
leftmatchContig = int(matchLocation[1])
aux = rightmatchAllele
rightmatchAllele = leftmatchAllele
leftmatchAllele = aux
else:
rightmatchContig = bestMatchContigLen - int(
matchLocation[1])
leftmatchContig = int(matchLocation[0])
###########################
# LOCUS ON THE CONTIG TIP #
###########################
# check if contig is smaller than the matched allele
if leftmatchContig < leftmatchAllele and rightmatchContig < rightmatchAllele:
# ~ resultsList.append('PLOTSC:-1')
perfectMatchIdAllele.append('LOTSC')
perfectMatchIdAllele2.append('LOTSC')
# ~ if not Reversed:
# ~ perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"+")
# ~ else:
# ~ perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"-")
verboseprint(match, contigname, geneFile,
leftmatchAllele, rightmatchAllele,
"Locus is bigger than the contig \n")
elif leftmatchContig < leftmatchAllele:
# ~ resultsList.append('PLOT3:-1')
perfectMatchIdAllele.append('PLOT3')
perfectMatchIdAllele2.append('PLOT3')
# ~ if not Reversed:
# ~ perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"+")
# ~ else:
# ~ perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"-")
verboseprint(
match, contigname, geneFile, leftmatchAllele,
rightmatchAllele,
"Locus is on the 3' tip of the contig \n")
elif rightmatchContig < rightmatchAllele:
# ~ resultsList.append('PLOT5:-1')
perfectMatchIdAllele.append('PLOT5')
perfectMatchIdAllele2.append('PLOT5')
# ~ if not Reversed:
# ~ perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"+")
# ~ else:
# ~ perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"-")
verboseprint(
match, contigname, geneFile, leftmatchAllele,
rightmatchAllele,
"Locus is on the 5' tip of the contig \n")
elif float(len(alleleStr)) > moda + (moda * 0.2):
verboseprint("Locus is larger than mode", moda,
alleleStr)
# ~ resultsList.append('ALM')
perfectMatchIdAllele.append('ALM')
perfectMatchIdAllele2.append('ALM')
# ~ if not Reversed:
# ~ perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"+")
# ~ else:
# ~ perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"-")
elif float(len(alleleStr)) < moda - (moda * 0.2):
verboseprint("Locus is smaller than mode", moda,
alleleStr)
# ~ resultsList.append('ASM')
perfectMatchIdAllele.append('ASM')
perfectMatchIdAllele2.append('ASM')
# ~ if not Reversed:
# ~ perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"+")
# ~ else:
# ~ perfectMatchIdAllele2.append(str(contigname)+"&"+str(matchLocation[0])+"-"+str(matchLocation[1])+"&"+"-")
else:
#######################
# ADD INFERRED ALLELE # # a new allele
#######################
wasContained = False
tagAuxC = 'S'
for alleleaux in fullAlleleList:
if alleleStr in alleleaux:
alleleName = fullAlleleNameList[
fullAlleleList.index(alleleaux)]
alleleMatchid = (alleleName.split("_"))[-1]
tagAuxC = 'CD' + alleleMatchid.rstrip()
resultsList.append([
(os.path.basename(genomeFile)),
str(alleleI + 1), tagAuxC
])
break
elif alleleaux in alleleStr:
alleleName = fullAlleleNameList[
fullAlleleList.index(alleleaux)]
alleleMatchid = (alleleName.split("_"))[-1]
tagAuxC = 'CS' + alleleMatchid.rstrip()
resultsList.append([
(os.path.basename(genomeFile)),
str(alleleI + 1), tagAuxC
])
break
if not wasContained:
tagAux = 'INF'
perfectMatchIdAllele.append(tagAux + "-" +
str(alleleI + 1))
if not Reversed:
perfectMatchIdAllele2.append(
str(contigname) + "&" +
str(matchLocation[0]) + "-" +
str(matchLocation[1]) + "&" + "+")
else:
perfectMatchIdAllele2.append(
str(contigname) + "&" +
str(matchLocation[0]) + "-" +
str(matchLocation[1]) + "&" + "-")
verboseprint("New allele! Adding allele " +
tagAux + str(alleleI + 1) +
" to the database\n")
# --- add the new allele to the gene fasta --- #
alleleI += 1
appendAllele = '>' + str((
((os.path.basename(geneFile)).split("."))[0]
).replace("_", "-")) + "_" + tagAuxC + "_" + (str(
os.path.basename(genomeFile))).replace(
"_", "-") + "_" + str(alleleI) + '\n'
fG = open(geneFile, 'a')
fG.write(appendAllele)
fG.write(alleleStr + '\n')
fG.close()
fullAlleleList.append(alleleStr)
fullAlleleNameList.append(appendAllele)
if bestmatch[1] >= int(bsrTresh) and float(
bestmatch[1]) < int(bsrTresh) + 0.1:
fG = open(shortgeneFile, 'a')
fG.write(appendAllele)
fG.write(alleleStr + '\n')
fG.close()
geneTransalatedPath2 = os.path.join(
basepath,
str(
os.path.basename(shortgeneFile) +
'_protein2.fasta'))
geneTransalatedPath = os.path.join(
basepath,
str(
os.path.basename(shortgeneFile) +
'_protein.fasta'))
with open(geneTransalatedPath2, 'w') as fG:
fG.write('>' + str(alleleI) + '\n' +
str(protSeq) + '\n')
with open(geneTransalatedPath, 'a') as fG:
fG.write('>' + str(alleleI) + '\n' +
str(protSeq) + '\n')
match = bestmatch[5]
# --- remake blast DB and recalculate the BSR for the locus --- #
alleleList.append(alleleStr)
listShortAllelesNames.append(appendAllele)
genefile2 = geneTransalatedPath2
Gene_Blast_DB_name2 = Create_Blastdb(
genefile2, 1, True)
verboseprint(
"Re-calculating BSR at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
allelescores, alleleList, listShortAllelesNames = reDogetBlastScoreRatios(
genefile2, basepath, alleleI, allelescores,
Gene_Blast_DB_name2, alleleList,
geneScorePickle, verbose, blastPath,
listShortAllelesNames)
verboseprint(
"Done Re-calculating BSR at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
except Exception as e:
print("some error occurred")
print(e)
print('Error on line {}'.format(sys.exc_info()[-1].tb_lineno))
perfectMatchIdAllele2.append("ERROR")
perfectMatchIdAllele.append("ERROR")
final = (resultsList, perfectMatchIdAllele)
verboseprint("Finished allele calling at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
filepath = os.path.join(temppath,
os.path.basename(geneFile) + "_result.txt")
filepath2 = os.path.join(temppath,
os.path.basename(geneFile) + "_result2.txt")
with open(filepath, 'wb') as f:
pickle.dump(final, f)
with open(filepath2, 'wb') as f:
pickle.dump(perfectMatchIdAllele2, f)
shutil.rmtree(basepath)
return True
def main():
parser = argparse.ArgumentParser(
description=
"Given an ffn file, recovers the genes that are not paralogs and have a size bigger than the g parameter provided"
)
parser.add_argument('-i',
nargs='?',
type=str,
help='ffn file',
required=True)
parser.add_argument('-g',
nargs='?',
type=int,
help='int minimum size',
required=True)
args = parser.parse_args()
genes = args.i
sizethresh = args.g
passSteps = False
#translate to protein and create new file
abspath = os.path.abspath(genes)
filename = os.path.basename(genes)
abspath = abspath.replace(filename, '')
proteinfile = os.path.join(abspath, 'proteins.fasta')
geneDict = {}
protDict = {}
orderedprotDict = collections.OrderedDict()
alreadyIn = []
totalgenes = 0
repeatedgenes = 0
smallgenes = 0
if not passSteps:
print "not passing steps"
with open(proteinfile, "wb") as f:
g_fp = HTSeq.FastaReader(genes)
totalgenes += 1
for gene in g_fp:
dnaseq = str(gene.seq)
protseq, x, y = translateSeq(dnaseq)
if len(protseq) > 1:
if str(protseq) in alreadyIn:
repeatedgenes += 1
elif len(str(protseq)) < 67:
smallgenes += 1
else:
alreadyIn.append(str(protseq))
protname = ">" + str(gene.name) + "\n"
f.write(protname + str(protseq) + "\n")
protDict[protname] = str(protseq)
geneDict[str(gene.name)] = gene.seq
else:
print gene.name
orderedprotList = []
orderedprotList = sorted(protDict.items(),
key=lambda x: len(x[1]),
reverse=True)
i = 0
while i < len(orderedprotList):
elem = orderedprotList[i]
orderedprotDict[elem[0]] = elem[1]
i += 1
#print orderedprotDict
print str(repeatedgenes) + " repeated genes out of " + str(totalgenes)
print str(smallgenes) + " small genes out of " + str(totalgenes)
print "protein file created"
# first step - remove genes contained in other genes or 100% equal genes
# list of results - the output of the function
resultsList = []
auxDict = {}
g_fp = HTSeq.FastaReader(proteinfile)
g = 0
j = 0
print "Checking if proteins are equal or substring of others..."
# for each gene from all the annotated genes - starting with an empty dictionary, only add a new gene if the "to be added gene" is not contained or equal to a gene already added to the dictionary
auxprot = []
for elem in orderedprotDict.items():
contained = False
prot = str(elem[1])
if any(prot in x for x in auxprot):
g += 1
contained = True
else:
auxDict[elem[1]] = elem[0]
auxprot.append(str(elem[1]))
print str(j) + " out of " + str(len(orderedprotDict))
j += 1
#print "____" +str(j)
print "%s genes are contained in other genes" % (g)
#overwrite the original file, obtaining a new file with unique genes
with open(proteinfile, "wb") as f:
allsequences = ''
for k, v in auxDict.iteritems():
allsequences += v + k + "\n"
f.write(allsequences)
else:
totalgenes = 0
smallgenes = 0
g_fp = HTSeq.FastaReader(genes)
totalgenes += 1
for gene in g_fp:
dnaseq = str(gene.seq)
protseq, x, y = translateSeq(dnaseq)
if len(protseq) > 1:
if str(protseq) in alreadyIn:
repeatedgenes += 1
#print gene.name + " already saved "
elif len(str(protseq)) < 67:
smallgenes += 1
else:
alreadyIn.append(str(protseq))
protname = ">" + str(gene.name) + "\n"
#print protseq
protDict[protname] = str(protseq)
geneDict[str(gene.name)] = gene.seq
else:
print gene.name
geneFile = os.path.abspath(proteinfile)
print proteinfile
Gene_Blast_DB_name = Create_Blastdb(geneFile, 1, True)
geneF = os.path.splitext(geneFile)[0]
blast_out_file = geneF + '.xml'
# ------------------------------ RUNNING BLAST ------------------------------ #
cline = NcbiblastpCommandline(query=geneFile,
db=Gene_Blast_DB_name,
evalue=0.001,
out=blast_out_file,
outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, geneFile)
toRemove = []
genesToKeep = []
log = ["removed\tcause\texplanation"]
for blast_record in blast_records:
allelename = blast_record.query
allelename = allelename.split(" ")
allelename = allelename[0]
alleleLength = len(geneDict[allelename])
try:
#if gene A is not on the toRemove list yet, add to genesToKeep list
if str(blast_record.query) not in toRemove:
genesToKeep.append(blast_record.query)
i = 0
#if first alignement is not against self, gene B is bigger than gene A and very simillar - remove gene A from genesToKeep and add gene B instead
if not str(blast_record.query) == str(
(blast_record.alignments[0]).hit_def):
genesToKeep.remove(str(blast_record.query))
toRemove.append(str(blast_record.query))
log.append(
str(blast_record.query) + "\t" +
str((blast_record.alignments[0]).hit_def) + "\t" +
"2 is first best match")
#if gene B is not on the toRemove list, add to genesToKeep list
if str(
(blast_record.alignments[0]).hit_def) not in toRemove:
genesToKeep.append(
str((blast_record.alignments[0]).hit_def))
raise
selfblastscore = (((blast_record.alignments[0]).hsps)[0]).score
while i < len(blast_record.alignments):
align = blast_record.alignments[i]
match = (align.hsps)[0]
scoreRatio = float(match.score) / float(selfblastscore)
alleleLength2 = len(geneDict[str(align.hit_def)])
#if good match and gene B not in toremove list
if (scoreRatio > 0.6 and
not str(align.hit_def) == str(blast_record.query)
and str(align.hit_def) not in toRemove):
#if gene B is bigger than gene A, keep bigger gene B
if alleleLength2 > alleleLength:
genesToKeep.append(str(align.hit_def))
genesToKeep.remove(str(blast_record.query))
toRemove.append(str(blast_record.query))
log.append(
str(blast_record.query) + "\t" +
str(align.hit_def) + "\t" +
"2 is bigger and bsr >0.6")
raise
#else add gene B to toremove list
elif str(align.hit_def) in genesToKeep:
genesToKeep.remove(str(align.hit_def))
toRemove.append(str(align.hit_def))
log.append(
str(align.hit_def) + "\t" +
str(blast_record.query) + "\t" +
"2 is bigger and bsr >0.6")
i += 1
#else gene A is on toRemove list, add all similar genes (not in genesToKeep) list to the toRemove list
else:
i = 0
selfblastscore = 0
for align in blast_record.alignments:
if not (str(align.hit_def) == str(blast_record.query)):
selfblastscore = ((align.hsps)[0]).score
print "gene " + str(
align.hit_def) + " is bigger than gene " + str(
blast_record.query)
raise
while i < len(blast_record.alignments):
align = blast_record.alignments[i]
match = (align.hsps)[0]
scoreRatio = float(match.score) / float(selfblastscore)
if align.hit_def not in genesToKeep and not str(
align.hit_def) == str(
blast_record.query) and scoreRatio > 0.6:
toRemove.append(align.hit_def)
log.append(
str(align.hit_def) + "\t" +
str(blast_record.query) + "\t" +
"2 was on the removed list and bsr >0.6")
else:
pass
i += 1
except Exception as e:
#print e
pass
with open("logfile.txt", "wb") as f:
for elem in log:
f.write(str(elem) + "\n")
genesToKeep = list(set(genesToKeep))
toRemove = list(set(toRemove))
s = set(toRemove)
notcommonToKeep = [x for x in genesToKeep if x not in s]
print len(toRemove)
print len(genesToKeep)
print len(notcommonToKeep)
pathfiles = os.path.dirname(geneFile)
pathfiles = pathfiles + "/"
g_fp = HTSeq.FastaReader(genes)
removedparalogs = 0
removedsize = 0
totalgenes = 0
rest = 0
concatenatedFile = ''
for contig in g_fp:
totalgenes += 1
name = contig.name + " " + contig.descr
name2 = contig.name
if name2 not in toRemove and name2 in genesToKeep:
if int(len(contig.seq)) > sizethresh:
namefile = contig.name
namefile = namefile.replace("|", "_")
with open(pathfiles + namefile + ".fasta", "wb") as f:
f.write(">1\n" + contig.seq + "\n")
rest += 1
concatenatedFile += ">" + namefile + "\n" + contig.seq + "\n"
else:
removedsize += 1
else:
removedparalogs += 1
print "%s genes are contained in other genes" % (g)
print "Removed %s same Locus genes" % str(removedparalogs)
print "Removed %s because of size " % str(removedsize)
print "%s Scheme genes " % str(rest)
print "total genes:" + str(totalgenes)
with open(pathfiles + "concatenated.fasta", "wb") as f:
f.write(concatenatedFile)
def main():
try:
input_file = sys.argv[1]
temppath = sys.argv[2]
except IndexError:
print "usage: list_pickle_obj"
argumentList = []
with open(input_file, 'rb') as f:
argumentList = pickle.load(f)
geneFile = argumentList[0]
genomesList = argumentList[1]
basepath = temppath + "/" + os.path.basename(geneFile)
if not os.path.exists(basepath + "/blastdbs/"):
os.makedirs(basepath + "/blastdbs/")
gene_fp = HTSeq.FastaReader(geneFile)
geneDict = {}
alleleI = 1
inverted = False
orderedAlleleNames = []
biggestAllelelen = 0
smallestAllelelen = 999999
for allele in gene_fp:
if allele.seq in geneDict:
print "\nWARNING: this file contains a repeated allele, it should be checked. Ignoring it now!\n", geneFile
else:
if len(allele.seq) > biggestAllelelen:
biggestAllelelen = len(allele.seq)
if len(allele.seq) < smallestAllelelen:
smallestAllelelen = len(allele.seq)
orderedAlleleNames.append(str(alleleI))
geneDict[allele.seq] = alleleI
alleleI += 1
# --- make 1st blast DB --- #
geneF = os.path.basename(geneFile)
blast_out_file = os.path.dirname(geneFile) + "/blastdbs/" + geneF + '.xml'
# list of results - the output of the function
i = 0
perfectMatchIdAllele = []
perfectMatchIdAllele2 = []
genomeDict = {}
genome = -1
resultsList = []
print genomesList
for genomeFile in genomesList:
print "_______________________________________________________"
print perfectMatchIdAllele
printinfo(genomeFile, geneFile)
#currentCDSDict = listOfCDSDicts[i]
g_fp = HTSeq.FastaReader(genomeFile)
for contig in g_fp:
sequence = str(contig.seq)
genomeDict[contig.name] = sequence
currentGenomeDict = genomeDict
genome += 1
print("Blasting alleles on genome at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
blast_out_file = os.path.join(
basepath, "blastdbs/" + os.path.basename(geneFile) + '_List.xml')
Gene_Blast_DB_name = os.path.join(
temppath,
str(os.path.basename(genomeFile)) + "/" +
str(os.path.basename(genomeFile)) + "_db")
cline = NcbiblastnCommandline(query=geneFile,
db=Gene_Blast_DB_name,
evalue=0.001,
out=blast_out_file,
outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, geneFile)
print("Blasted alleles on genome at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
# ------ DETERMINING BEST MATCH ------ #
bestMatch = ['', '', 0]
bestMatchContig = ''
bestMatchContigLen = ''
bestalignlen = 0
perfectMatch = False
bmAlleleLen2 = 0
bmAllele = ''
for blast_record in blast_records:
if perfectMatch == True:
break
try:
hspC = blast_record.alignments[0]
if bestMatch[0] == '' and bestMatch[1] == '':
bestMatch[0] = blast_record.query
bestMatch[1] = hspC
except IndexError:
continue
# --- the contig tag is used in the progigal function --- #
contigTag = blast_record.query
# --- brute force parsing of the contig tag - better solution is advisable --- #
j = 0
for l in contigTag:
if l == ' ':
break
j += 1
contigTag = contigTag[:j]
contigLen = blast_record.query_letters
# --- iterating over all the results to determine the best match --- #
for alignment in blast_record.alignments:
contigTag = alignment.hit_def
contigTag = (contigTag.split(" "))[0]
index = orderedAlleleNames.index(
str(blast_record.query_id).split("_")[1])
for k, v in geneDict.iteritems():
if v == index + 1:
bmAlleleLen2 = len(k)
if perfectMatch:
break
for match in alignment.hsps:
scoreRatio = float(match.score) / float(bmAlleleLen2)
#if #identities is the same as the length of the allele and it has no gaps or N's
if (int(match.identities) == int(bmAlleleLen2)
and int(match.identities) == int(len(match.query))
and "N" not in match.sbjct
and "K" not in match.sbjct
and "Y" not in match.sbjct
and "R" not in match.sbjct):
index = orderedAlleleNames.index(
str(blast_record.query_id).split("_")[1])
for seq, alleleid in geneDict.iteritems():
if alleleid == index + 1:
bmAllele = seq
break
bmAlleleLen = len(bmAllele)
lenratio = float(len(match.sbjct)) / float(bmAlleleLen)
bestMatch = [
blast_record.query, match, scoreRatio,
blast_record.query_id, lenratio, bmAlleleLen
]
bestMatchContig = contigTag
perfectMatch = True
index = orderedAlleleNames.index(
str(blast_record.query_id).split("_")[1])
bmAlleleLen = len(geneDict.keys()[index])
break
#chose the match with the best score ratio (score/length of allele)
elif scoreRatio > bestMatch[2]:
index = orderedAlleleNames.index(
str(blast_record.query_id).split("_")[1])
for seq, alleleid in geneDict.iteritems():
if alleleid == index + 1:
bmAllele = seq
break
bmAlleleLen = len(bmAllele)
lenratio = float(len(match.sbjct)) / float(bmAlleleLen)
bestMatch = [
blast_record.query, match, scoreRatio,
blast_record.query_id, lenratio, bmAlleleLen
]
bestMatchContig = contigTag
bestMatchContigLen = len(currentGenomeDict[contigTag])
print contigTag
bestalignlen = alignment.length
if perfectMatch == True:
break
# ---------- ALLELE CALLING AFTER DETERMINING BEST MATCH ---------- #
print("Finished choosing best match at : " +
time.strftime("%H:%M:%S-%d/%m/%Y"))
try:
match = bestMatch[1]
bestMatchStart = match.sbjct_start
bestMatchEnd = match.sbjct_end
if match.query_start > match.query_end:
bestMatchEnd = match.sbjct_start
bestMatchStart = match.sbjct_end
print match
geneLen = bestMatch[5]
alleleStr = match.sbjct
nIdentities = match.identities
idPercent = float(nIdentities) / float(geneLen)
scoreRatio = bestMatch[2]
lenRatio = bestMatch[4]
except:
#if no best match was found
###################
# LOCUS NOT FOUND #
###################
perfectMatchIdAllele.append('LNF')
perfectMatchIdAllele2.append('LNF')
print "Locus not found, no matches \n"
continue
print "is perfect match true?" + str(perfectMatch)
if perfectMatch is True:
#if a perfect match was found
try:
alleleNumber = geneDict[alleleStr]
except:
alleleStr = reverseComplement(alleleStr)
alleleNumber = geneDict[alleleStr]
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
if "_" in bestMatch[3]:
a = bestMatch[3].split("_")
perfectMatchIdAllele.append(a[1])
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(bestMatchStart) + "-" +
str(bestMatchEnd) + "&" + "+")
else:
perfectMatchIdAllele.append(bestMatch[3])
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(bestMatchStart) + "-" +
str(bestMatchEnd) + "&" + "+")
printinfo(genomeFile, geneFile)
print "Exact match \n"
continue
else:
#if a best match was found but it's not an exact match
###########################
# LOCUS ON THE CONTIG TIP #
###########################
print geneLen
if bestMatchContigLen <= geneLen:
perfectMatchIdAllele.append('LOTSC')
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(bestMatchStart) + "-" +
str(bestMatchEnd) + "&" + "+")
printinfo(genomeFile, geneFile)
print "Locus is bigger than the contig \n"
elif (match.sbjct_start == 1 and len(match.query) < geneLen) or (
match.sbjct_start == bestMatchContigLen
and len(match.query) < bestMatchContigLen
and match.sbjct_start > match.sbjct_end):
perfectMatchIdAllele.append('LOT5')
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(bestMatchStart) + "-" +
str(bestMatchEnd) + "&" + "+")
printinfo(genomeFile, geneFile)
print "Locus is on the 5' tip of the contig \n"
elif (match.sbjct_end == 1 and len(match.query) < geneLen
and match.sbjct_start > match.sbjct_end) or (
match.sbjct_end == bestMatchContigLen
and len(match.query) < bestMatchContigLen):
perfectMatchIdAllele.append('LOT3')
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(bestMatchStart) + "-" +
str(bestMatchEnd) + "&" + "+")
printinfo(genomeFile, geneFile)
print "Locus is on the 3' tip of the contig \n"
elif 'N' in alleleStr or "K" in alleleStr or "R" in alleleStr or "Y" in alleleStr:
#####################
# ALLELE NOT FOUND # # N base found!
#####################
geneFile2 = os.path.splitext(geneFile)[0] + "LNFN.fasta"
with open(geneFile2, 'a') as f:
f.write(">" + (str(os.path.basename(genomeFile))) + "|" +
(str(os.path.basename(geneFile))) + "\n")
f.write((alleleStr) + "\n")
perfectMatchIdAllele.append('LNFN')
perfectMatchIdAllele2.append('LNFN')
printinfo(genomeFile, geneFile)
print "LNFN, contains strange (N,K,R) bases! \n"
else:
print "new allele?"
#removing gaps
alleleStr = alleleStr.replace('-', '')
lenExtraThresh = int(biggestAllelelen * 0.2)
#else: #check if best match without gaps are contained inside an already defined allele
isContainedDefinedAllele = False
definedAllele = ''
definedAlleleName = ''
for k in geneDict.keys():
if alleleStr in k:
definedAllele = k
isContainedDefinedAllele = True
definedAlleleName = geneDict.get(k)
break
print "is contained? " + str(isContainedDefinedAllele)
print idPercent
print geneLen
print lenExtraThresh
print lenRatio
if isContainedDefinedAllele and int(len(match.sbjct)) <= int(
len(definedAllele)) + lenExtraThresh and int(
len(match.sbjct)) >= int(
len(definedAllele)) - lenExtraThresh:
#allele without gaps is contained in a defined allele
#best match with gaps has same size +1/-1 base as the defined allele
isnewallele = False
if int(len(alleleStr)) == int(
len(definedAllele)
): # if match without gaps has same size as the defined allele
tagAux = 'NA?:'
printinfo(genomeFile, geneFile)
perfectMatchIdAllele.append("NA?-" + str(alleleI))
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(bestMatchStart) +
"-" + str(bestMatchEnd) + "&" + "+")
isnewallele = True
elif int(len(alleleStr)) == int(
len(definedAllele)
) - 1: # if match without gaps has minus one base than the defined allele
tagAux = 'NA2:'
printinfo(genomeFile, geneFile)
perfectMatchIdAllele.append("NA2-" + str(alleleI))
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(bestMatchStart) +
"-" + str(bestMatchEnd) + "&" + "+")
isnewallele = True
else:
extraleft = 0
extraright = 0
tS = 0
tE = 0
handle = open(genomeFile, "rU")
record_dict = SeqIO.to_dict(
SeqIO.parse(handle, "fasta"))
handle.close()
record = record_dict[bestMatchContig]
# if match without gaps has more than one base missing comparing to the defined allele
if (1 < int(match.query_start)
and 1 < int(match.query_end)):
if match.query_start > match.query_end:
extraleft = match.query_end - 1
else:
extraleft = match.query_start - 1
print extraleft, extraright
if (
int(geneLen) > int(match.query_start)
and int(geneLen) > int(match.query_end)
): # if 3' tip bases of the allele are missing on the match
if match.query_start > match.query_end:
extraright = geneLen - match.query_start
else:
extraright = geneLen - match.query_end
print extraleft, extraright
if match.sbjct_start > match.sbjct_end:
tE = match.sbjct_start + extraleft
tS = match.sbjct_end - extraright - 1
alleleStr = str(record.seq[tS:tE])
alleleStr = reverseComplement(alleleStr)
else:
tS = match.sbjct_start - extraleft - 1
tE = match.sbjct_end + extraright
alleleStr = str(record.seq[tS:tE])
print tS
print tE
print "allele is:"
print alleleStr
if tE > bestMatchContigLen:
perfectMatchIdAllele.append('LOT3B')
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(tS) + "-" +
str(bestMatchContigLen) + "&" + "+")
printinfo(genomeFile, geneFile)
print "Locus is on the 3B' tip of the contig \n"
elif tS < 0:
perfectMatchIdAllele.append('LOT5B')
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(0) + "-" +
str(tE) + "&" + "+")
printinfo(genomeFile, geneFile)
print "Locus is on the 5B' tip of the contig \n"
else:
tagAux = 'NA2:'
printinfo(genomeFile, geneFile)
perfectMatchIdAllele.append("NA2-" + str(alleleI))
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(tS) + "-" +
str(tE) + "&" + "+")
isnewallele = True
if isnewallele:
print "New allele found! Adding allele " + tagAux + str(
alleleI) + " to the database"
geneDict[alleleStr] = alleleI
orderedAlleleNames.append(str(alleleI))
# --- add the new allele to the gene fasta --- #
fG = open(geneFile, 'a')
fG.write('>allele_' + str(alleleI) + '_' +
tagAux[:-1] + '_' +
str(os.path.basename(genomeFile)) + '\n')
fG.write(alleleStr + '\n')
fG.close()
alleleI += 1
#if best match is not contained in an already defined allele, check if it has similar size with the match allele and has 0.8 similarity
elif not isContainedDefinedAllele and idPercent >= 0.8 and int(
len(match.sbjct)
) <= int(geneLen) + lenExtraThresh and int(len(
match.sbjct)) >= int(geneLen) - lenExtraThresh:
#best match with gaps has 80% identity
#best match with gaps is the same size or +1/-1 as the defined allele
ratio = float(len(alleleStr)) / float(geneLen)
if ratio >= 0.8 and ratio <= 1.2: # if match without gaps has same size as the best match allele and 80%similarity
tagAux = ''
extraleft = 0
extraright = 0
tS = 0
tE = 0
handle = open(genomeFile, "rU")
record_dict = SeqIO.to_dict(
SeqIO.parse(handle, "fasta"))
handle.close()
record = record_dict[bestMatchContig]
#if len(match.sbjct)<geneLen and "-" not in match.sbjct: #if the allele is not fully covered against the match, compensate the tips
try:
print match
if (1 < int(match.query_start)
and 1 < int(match.query_end)):
if match.query_start > match.query_end:
extraleft = match.query_end - 1
else:
extraleft = match.query_start - 1
print extraleft, extraright
if (
int(geneLen) > int(match.query_start)
and int(geneLen) > int(match.query_end)
): # if 3' tip bases of the allele are missing on the match
if match.query_start > match.query_end:
extraright = geneLen - match.query_start
else:
extraright = geneLen - match.query_end
print extraleft, extraright
if match.sbjct_start > match.sbjct_end:
tE = match.sbjct_start + extraleft
tS = match.sbjct_end - extraright - 1
alleleStr = str(record.seq[tS:tE])
alleleStr = reverseComplement(alleleStr)
else:
tS = match.sbjct_start - extraleft - 1
tE = match.sbjct_end + extraright
alleleStr = str(record.seq[tS:tE])
print tS
print tE
print "allele is:"
print alleleStr
if tE > bestMatchContigLen:
perfectMatchIdAllele.append('LOT3C')
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(tS) +
"-" + str(bestMatchContigLen) + "&" + "+")
printinfo(genomeFile, geneFile)
print "Locus is on the 3C' tip of the contig \n"
elif tS < 0:
perfectMatchIdAllele.append('LOT5C')
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(0) + "-" +
str(tE) + "&" + "+")
printinfo(genomeFile, geneFile)
print "Locus is on the 5C' tip of the contig \n"
else:
tagAux = 'NA3:'
printinfo(genomeFile, geneFile)
perfectMatchIdAllele.append("NA3-" +
str(alleleI))
perfectMatchIdAllele2.append(
str(bestMatchContig) + "&" + str(tS) +
"-" + str(tE) + "&" + "+")
print "New allele found! Adding allele " + tagAux + str(
alleleI) + " to the database"
geneDict[alleleStr] = alleleI
orderedAlleleNames.append(str(alleleI))
# --- add the new allele to the gene fasta --- #
fG = open(geneFile, 'a')
fG.write('>allele_' + str(alleleI) + '_' +
tagAux[:-1] + '_' +
str(os.path.basename(genomeFile)) +
'\n')
fG.write(alleleStr + '\n')
fG.close()
alleleI += 1
except Exception as e:
##################
# LNF #
##################
print e
geneFile2 = os.path.splitext(
geneFile)[0] + "LNF3.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">" +
(str(os.path.basename(genomeFile))) +
"|" +
(str(os.path.basename(geneFile))) +
" | " + str(bestMatchContig) + "\n")
f.write((alleleStr) + "\n")
f.write(">Allele\n")
f.write((bmAllele) + "\n")
printinfo(genomeFile, geneFile)
perfectMatchIdAllele.append("LNF3")
perfectMatchIdAllele2.append("LNF3")
print "No allele found"
else:
##################
# LNF #
##################
geneFile2 = os.path.splitext(
geneFile)[0] + "LNF4.fasta"
print geneFile2
with open(geneFile2, 'a') as f:
f.write(">" + (str(os.path.basename(genomeFile))) +
"|" + (str(os.path.basename(geneFile))) +
" | " + str(bestMatchContig) + "\n")
f.write((alleleStr) + "\n")
f.write(">Allele\n")
f.write((bmAllele) + "\n")
printinfo(genomeFile, geneFile)
perfectMatchIdAllele.append("LNF4")
perfectMatchIdAllele2.append("LNF4")
print "No allele found"
elif isContainedDefinedAllele:
####################
# UNDEFINED ALLELE # # it is contained in another allele
####################
alleleStr = match.query
perfectMatchIdAllele.append("undefined allele")
perfectMatchIdAllele2.append("undefined allele")
printinfo(genomeFile, geneFile)
print "Undefined allele \n"
geneFile2 = os.path.splitext(
geneFile)[0] + "undefined.fasta"
print geneFile2
elif lenRatio < 0.5:
###############
# SMALL MATCH #
###############
perfectMatchIdAllele.append('small match')
perfectMatchIdAllele2.append('small match')
printinfo(genomeFile, geneFile)
print "lower than 50% match \n"
elif lenRatio < 0.8 and idPercent < 0.5:
#####################
# INCOMPLETE ALLELE # # it was not possible to extend it to at least 80% of the length of the gene
#####################
perfectMatchIdAllele.append('allele incomplete')
perfectMatchIdAllele2.append('allele incomplete')
printinfo(genomeFile, geneFile)
print "Incomplete allele\n"
else:
##################
# LNF #
##################
printinfo(genomeFile, geneFile)
perfectMatchIdAllele.append("LNF5")
perfectMatchIdAllele2.append("LNF5")
print "Locus not found"
final = (resultsList, perfectMatchIdAllele)
print("Finished allele calling at : " + time.strftime("%H:%M:%S-%d/%m/%Y"))
filepath = os.path.join(temppath,
os.path.basename(geneFile) + "_result.txt")
filepath2 = os.path.join(temppath,
os.path.basename(geneFile) + "_result2.txt")
with open(filepath, 'wb') as f:
pickle.dump(final, f)
with open(filepath2, 'wb') as f:
pickle.dump(perfectMatchIdAllele2, f)
return True
def callAlleles(argumentList):
geneFile = argumentList[0]
genomesList = argumentList[1]
listOfCDSDicts = argumentList[2]
listOfGenomesDict = argumentList[3]
gene_fp = HTSeq.FastaReader(geneFile)
geneDict = {}
alleleI = 0
for allele in gene_fp:
if allele.seq in geneDict:
print "\nWARNING: this file contains a repeated allele, it should be checked. Ignoring it now!\n", geneFile
else:
geneDict[ allele.seq ] = alleleI
alleleI += 1
# --- make 1st blast DB --- #
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 )
geneF = os.path.splitext( geneFile )[0]
blast_out_file = geneF + '.xml'
# list of results - the output of the function
resultsList = []
i = 0
for genomeFile in genomesList:
currentCDSDict = listOfCDSDicts[i]
currentGenomeDict = listOfGenomesDict[i]
i+=1 # it has to be incremented here
if genomeFile[-1] == '\n':
genomeFile = genomeFile[:-1]
# ------------------------------ RUNNING BLAST ------------------------------ #
cline = NcbiblastnCommandline(query=genomeFile, db=Gene_Blast_DB_name, evalue=0.001, out=blast_out_file, outfmt=5)
blast_records = runBlastParser(cline, blast_out_file, genomeFile)
# ------ DETERMINING BEST MATCH ------ #
# bestMatch = ['rec.query','hsp', lenRatio]
bestMatch = ['','', 0]
for blast_record in blast_records:
# --- the LNF cases are now called outside de loop --- #
try:
hspC = blast_record.alignments[0]
if bestMatch[0] == '' and bestMatch[1] == '':
bestMatch[0] = blast_record.query
bestMatch[1] = hspC
except IndexError:
continue
# --- the contig tag is used in the progigal function --- #
contigTag = blast_record.query
# --- brute force parsing of the contig tag - better solution is advisable --- #
j=0
for l in contigTag:
if l == ' ':
break
j+=1
contigTag = contigTag[:j]
contigLen = blast_record.query_letters
# --- iterating over all the results to determine the best match --- #
for alignment in blast_record.alignments:
for match in alignment.hsps:
lenRatio = float(len( match.query )) / float( len(match.sbjct) )
if lenRatio > bestMatch[2]:
bestMatch = [blast_record.query, match, lenRatio]
# ---------- ALLELE CALLING AFTER DETERMINING BEST MATCH ---------- #
###################
# LOCUS NOT FOUND #
###################
if bestMatch[0] == '':
resultsList.append('LNF:-1') # append result to the list of results
continue
match = bestMatch[1]
geneLen = len(match.sbjct)
alleleStr = match.query
nIdentities = match.identities
idPercent = float(nIdentities) / float(geneLen)
lenRatio = bestMatch[2]
###########################
# LOCUS ON THE CONTIG TIP #
###########################
if contigLen <= match.query_start or contigLen <= match.query_end:
resultsList.append('LOT:-1')
###############
# SMALL MATCH #
###############
elif lenRatio < 0.5:
resultsList.append('SAC:-1') # don't know what 'SAC' stands for
else:
# ------------------------------------------------------------------------------------------------------- #
# #
# USING PRODIGAL TO TRY TO EXTEND CDS #
# #
# ------------------------------------------------------------------------------------------------------- #
extended, strCDS = extendCDS(contigTag, currentCDSDict, match.sbjct_start, match.sbjct_end, currentGenomeDict)
# --- if it was possible to extend it using prodigal --- #
if extended and ( ( len(strCDS) * lenRatio ) >= geneLen): # and idPercent > 0.8 and ( len(strCDS) / geneLen) > 0.8:
alleleStr = strCDS
lenRatio = float(len(strCDS)) / float(geneLen)
# --- removing gaps '-' --- #
alleleStr = alleleStr.replace('-', '')
# --- continuing the allele calling --- #
if lenRatio < 0.8 and idPercent < 0.5:
#####################
# INCOMPLETE ALLELE # # it was not possible to extend it to at least 80% of the length of the gene
#####################
resultsList.append('INC:-1')
else:
# --- it might be needed to obtain the reverse complement of the allele string --- #
if match.sbjct_start > match.sbjct_end:
alleleStr = reverseComplement(alleleStr)
if alleleStr in geneDict:
alleleNumber = geneDict[ alleleStr ]
################################################
# EXACT MATCH --- MATCH == GENE --- GENE FOUND #
################################################
resultsList.append('EXC:' + str(alleleNumber) )
else:
isUndefined = False
for k in geneDict.keys():
if alleleStr in k:
isUndefined = True
break
if isUndefined:
####################
# UNDEFINED ALLELE # # it is contained in another allele
####################
resultsList.append('UND:-1')
else:
if not extended and idPercent > 0.8:
##################
# ADD NEW ALLELE #
##################
tagAux = 'NA:'
else:
#######################
# ADD INFERRED ALLELE # # a new allele that was extended with prodigal
#######################
tagAux = 'INF:'
resultsList.append( tagAux + str(alleleI) )
geneDict[ alleleStr ] = alleleI
alleleI += 1
# --- add the new allele to the gene fasta --- #
fG = open( geneFile, 'a' )
fG.write('>allele_' + str(alleleI) + '_' + tagAux[:-1] + '\n')
fG.write( alleleStr + '\n')
fG.close()
# --- remake blast DB --- #
Gene_Blast_DB_name = Create_Blastdb( geneFile, 1 )
return resultsList