def searchAndSave(searchString, filename):
""" convenince function to search database and save
results to a FASTA file. If a file of that name
already exists then the whole process is skipped
searchString - this is what is being searched for
filename - name of the file where to save the results
"""
if not(path.isfile(filename)):
sequences = []
""" set of ids returned as part of the protein """
ids = seq.searchSequences(searchString)
print "Processing ", len(ids), " sequences ..."
""" iterate over the ids and for each fetch the record
from the database and append it to ex5_sequences
"""
for seq_id in ids:
print "Fetching sequence: ", seq_id
sequences.append(seq.getSequence(seq_id))
""" save the completed list of sequences to a FASTA file """
seq.writeFastaFile(filename, sequences)
else:
print filename, " exists. skipping."
def allMotifs_fa(args):
#check hoow many cols, check if all of them has a value
#make a FASTA and make a CSV
for i in range (len(args.input)):
c=0
fasta = {}
name = (args.input[i])
name = name.split('.')[0]
name = name + '_reduced.fa'
with open(args.input[i], newline='') as f:
reader = csv.reader(f)
header = next(reader)
for row in reader:
isEmpty = False
for i in range(1, len(header)-1):
if row[i] == "":
isEmpty = True
break
if isEmpty == False:
fasta[row[0]] = row[len(header)-1]
seq_list = [sequence.Sequence(sequence=seq, name=seqname) for seqname, seq in fasta.items()]
sequence.writeFastaFile(name, seq_list)
c+=1
print(str(len(seq_list)) + " sequences kept after applying the requirements for " + name)
def read(args):
outputfile = output(args)
orig_dict = {}
if '.csv' in args.input:
print("this is a CSV file")
outputfile = outputfile + '.fa'
with open(args.input, newline='') as f:
reader = csv.reader(f)
for row in reader:
orig_dict[row[0]] = row[1]
seq_list = [
sequence.Sequence(sequence=seq, name=seqname)
for seqname, seq in orig_dict.items()
]
sequence.writeFastaFile(outputfile, seq_list)
elif '.tab' in args.input or '.tsv' in args.input:
print("this is a TAB/TSV file")
outputfile = outputfile + '.fa'
with open(args.input) as tsv:
for line in csv.reader(tsv, dialect="excel-tab"):
orig_dict[line[0]] = line[1]
seq_list = [
sequence.Sequence(sequence=seq, name=seqname)
for seqname, seq in orig_dict.items()
]
sequence.writeFastaFile(outputfile, seq_list)
elif '.fa' in args.input or '.fasta' in args.input:
print("this is a FASTA file")
outputfile = outputfile + '.csv'
db100 = sequence.readFastaFile(args.input,
sequence.Protein_Alphabet,
ignore=True,
parse_defline=False)
with open(outputfile, 'w', newline='') as f:
fieldnames = ['Name', 'Sequence']
thewriter = csv.DictWriter(f, fieldnames=fieldnames)
thewriter.writeheader()
for seq in db100:
s = ''.join(seq.sequence)
thewriter.writerow({'Name': seq.name, 'Sequence': s})
def posEqual_fa(args):
fasta = {}
name = (args.input[0])
name = name.split('.')[0]
csvFile = name + '_reduced.csv'
seqCol = 0
with open(csvFile, newline='') as f:
reader = csv.reader(f)
header = next(reader)
seqCol = len(header)-1
for row in reader:
fasta[row[0]] = row[len(header)-1]
seq_list = [sequence.Sequence(sequence=seq, name=seqname) for seqname, seq in fasta.items()]
sequence.writeFastaFile(name+'.fa', seq_list)
def main():
parser = argparse.ArgumentParser()
parser.add_argument("-i",
"--input",
help="FASTA file to query from",
required=True)
parser.add_argument("-q",
"--query",
help="Query FASTA file",
required=True)
parser.add_argument("-db",
"--database",
help="Database output file name",
required=True)
parser.add_argument("-r",
"--reference",
help="Reference database ",
default="uniprotkb")
parser.add_argument("-o",
"--output",
help="Output path",
default="matchmyseqs")
args = parser.parse_args()
seqDict = {}
tier1seq = ''
representative = ''
fasta = {}
seqsforCSV = {}
progress = 0
tier1 = {}
tier1_annots = {
} # annotations that we want to include in the final dataset
os.system('makeblastdb -dbtype prot -in ' + args.input + ' -out ' +
args.database)
db = sequence.readFastaFile(args.input,
sequence.Protein_Alphabet,
ignore=True,
parse_defline=False)
db_map = {} # map from "long" name to actual entry
db_map_short = {} # map from "short" name to entry
for s in db:
db_map[s.name] = s
db_map_short[sequence.parseDefline(s.name)[0]] = s
print("Database size is " + str(len(db_map)))
print(
"Blast started, this might take a bit depending on your dataset size")
os.system("blastp -db " + args.database +
" -outfmt 3 -num_descriptions 1 -num_alignments 0 -query " +
args.query + " -out query.txt")
if args.reference == 'uniprotkb':
os.system(
"grep -e \"^[st][pr]|\" query.txt | cut -d\' \' -f1 > UniProt_query.tab"
)
# Extract the resulting sequence identifiers
repSeqNames = set([])
f = open('UniProt_query.tab', 'rt')
for row in f:
repSeqNames.add(sequence.parseDefline(row.strip())[0])
f.close()
print(str(len(repSeqNames)),
" representative sequences have been found")
#Annot the representative sequences
notfound = []
for name in repSeqNames:
if name in db_map_short:
s = db_map_short[name]
seqsforCSV[s.name] = "".join(s)
else:
notfound.append(name)
print('Matched',
len(repSeqNames) - len(notfound), 'of', len(repSeqNames))
with open("query.txt", newline='') as f:
reader = csv.reader(f)
for row in reader:
if len(row) > 0 and row[0].startswith('Query'):
querySeq = (str(row).split("=")[1][:-2].strip())
elif len(row) > 0 and (row[0].startswith('tr|')
or row[0].startswith('sp|')):
representative = (str(row).split(" ")[0][2:].strip())
seqDict[querySeq] = representative
elif args.reference == 'refseq':
grab = False
repSeqNames = set([])
with open("query.txt", newline='') as f:
reader = csv.reader(f)
for row in reader:
if len(row) > 0 and row[0].startswith('Query'):
querySeq = (str(
row[0]).split("=")[1][:-2].strip().split(" ")[0])
elif len(row) > 0 and row[0].startswith('Sequences'):
grab = True
continue
elif grab == True:
if len(row) > 0 and not row[0].strip() == "":
representative = (row[0].split('.')[0] + "." +
row[0].split('.')[1].split(" ")[0])
repSeqNames.add(representative)
seqDict[querySeq] = representative
grab = False
#print(len(repSeqNames))
notfound = []
for name in repSeqNames:
if name in db_map_short:
s = db_map_short[name]
seqsforCSV[s.name] = "".join(s)
else:
notfound.append(name)
print('Matched',
len(repSeqNames) - len(notfound), 'of', len(repSeqNames))
print(len(repSeqNames),
" representative sequences found for " + args.query)
# done25 = False
# done50 = False
# done75 = False
# for s,rep in seqDict.items():
# total = (len(seqDict))
# seq = (sequence.getSequence(rep,'uniprotkb'))
# seqsforCSV[rep] = str(seq).split(":")[1].strip()
# elem = rep + str(seq)
# progress+=1
# if (progress/total)*100 > 25 and not done25:
# print("25% done")
# done25 = True
# elif (progress/total)*100 > 50 and not done50:
# print("50% done")
# done50 = True
# elif (progress/total)*100 > 75 and not done75:
# print("75% done")
# done75 = True
faOut = args.output + '.fa'
seq_list = [
sequence.Sequence(sequence=seq, name=seqname)
for seqname, seq in seqsforCSV.items()
]
sequence.writeFastaFile(faOut, seq_list)
csvOut = args.output + '.csv'
with open(csvOut, 'w', newline='') as f:
fieldnames = ['Name', 'Representative', 'Sequence']
thewriter = csv.DictWriter(f, fieldnames=fieldnames)
thewriter.writeheader()
for given, rep in seqDict.items():
thewriter.writerow({
'Name': given,
'Representative': rep,
'Sequence': seqsforCSV[rep]
})
def main():
parser = argparse.ArgumentParser()
parser.add_argument("-i",
"--input",
help="Input FASTA file",
required=True)
parser.add_argument("-db",
"--database",
help="Database output file name",
required=True)
parser.add_argument("-r",
"--redundancy",
nargs='*',
help="List of redundancy levels",
default=[90, 80, 70])
parser.add_argument("-t1", "--tier1", help="User's Tier1 sequences")
parser.add_argument("-t2", "--tier2", help="User's Tier2 sequences")
parser.add_argument("-ml",
"--maxlength",
help="Max length that the sequence can be",
default=800)
parser.add_argument("-e",
"--eval",
nargs='*',
help="List of evalues",
default=[1e-100, 1e-75, 1e-50, 1e-20, 1e-10, 1e-5])
args = parser.parse_args()
tier2 = {}
tier2_short = {}
tier2_annots = {
} # annotations that we want to include in the final dataset
if args.tier2:
print("tier2 sequences have been provided")
if '.fa' in args.tier2 or '.fasta' in args.tier2:
print("tier2 sequences are FASTA file")
tier2db = sequence.readFastaFile(args.tier2,
sequence.Protein_Alphabet,
ignore=True,
parse_defline=False)
print(str(len(tier2_list)) + " sequences in tier2")
tier2_list = {} # map from "long" name to actual entry
tier2_map_short = {} # map from "short" name to entry
for s in tier2db:
tier2_list[s.name] = s
tier2_map_short[sequence.parseDefline(s.name)[0]] = s
else:
print("Please provide FASTA file for tier-2")
if args.tier1:
tier1 = {}
tier1_annots = {
} # annotations that we want to include in the final dataset
print("Tier-1 sequences have been provided")
if '.fa' in args.tier1 or '.fasta' in args.tier1:
print("Tier-1 sequences are provided as a FASTA file")
tier1db = sequence.readFastaFile(args.tier1,
sequence.Protein_Alphabet,
ignore=True,
parse_defline=False)
tier1_list = {}
for s in tier1db:
tier1_list[s.name] = "".join(s.sequence)
print("Tier-1 has " + str(len(tier1_list)) + " sequences")
else:
print("Please provide FASTA file for tier-1")
db100 = sequence.readFastaFile(args.input,
sequence.Protein_Alphabet,
ignore=True,
parse_defline=False)
db100_map = {} # map from "long" name to actual entry
db100_map_short = {} # map from "short" name to entry
for s in db100:
db100_map[s.name] = s
db100_map_short[sequence.parseDefline(s.name)[0]] = s
print("Database has " + str(len(db100_map)) + " sequences")
for rr in args.redundancy:
rs = str(rr)
os.system('cd-hit -i ' + args.input + ' -c 0.' + rs + ' -T 5 -o db' +
rs + ' -d 0')
selected = {}
for rr in args.redundancy:
selected[rr] = []
filename = 'db' + str(rr) + '.clstr'
clusters = readCDHIT(filename)
for c in clusters:
picked_one = False
shortest = None
reviewed = None
for name in clusters[c]:
if name in db100_map:
seq = db100_map[name]
if shortest:
if len(seq) < len(shortest) and not disqualified(
seq, args):
shortest = seq
elif not disqualified(seq, args):
shortest = seq
if seq.name.startswith('sp|') and not disqualified(
seq, args):
reviewed = seq
if name in tier1_list:
#print("this one orig" + str(seq))
selected[rr].append(seq)
picked_one = True
else:
pass
#print('Did not find', name)
# If no Tier-1, prefer "reviewed", then shortest length
if not picked_one and reviewed:
selected[rr].append(reviewed)
elif not picked_one and shortest:
selected[rr].append(shortest)
for rr in args.redundancy:
filename = 'db' + str(rr) + '.fa'
sequence.writeFastaFile(filename, selected[rr])
for rr in args.redundancy:
os.system('makeblastdb -dbtype prot -in db' + str(rr) +
'.fa -out db-' + str(rr))
# for rr in args.redundancy:
# for evalue in args.evalue:
# result_file = "dataset-" + str(rr) + '-'+ str(evalue)
# cmd1 = "blastp -db db-" + str(rr) + " -outfmt 3 -num_descriptions 20000 -num_alignments 0 -num_threads 5 -query " + args.tier1 + " -out " + result_file + ".txt -evalue " + str(evalue)
# print(cmd1)
# os.system(cmd1)
grab = False
for rr in args.redundancy:
for evalue in args.eval:
c = 0
tpsIdentifier = set([])
seqs = []
result_file = "dataset-" + str(rr) + '-' + str(evalue)
f = open(result_file + '.txt', 'rt')
for row in f:
if row.startswith('Sequences'):
grab = True
continue
if grab == True:
if row.startswith('Lambda'):
grab = False
if not row.strip() == "":
identifier = row.split(' ')[0]
if identifier != "Lambda":
tpsIdentifier.add(identifier)
for name in tpsIdentifier:
try:
seq = db100_map[name]
info = ''
seqs.append(
sequence.Sequence(seq.sequence, seq.alphabet, seq.name,
info))
except:
pass
sequence.writeFastaFile(result_file + ".fa", seqs)
print(result_file + " has " + str(len(seqs)) + "sequences")
print('Done')
totalSeqCount = []
c = 0
for evalue in args.eval:
for rr in args.redundancy:
output = []
ev = str(evalue)
ev = ev[1:]
red = str(rr)
result_file = "dataset-" + str(rr) + '-' + str(evalue)
a = sequence.readFastaFile(result_file + '.fa',
sequence.Protein_Alphabet,
ignore=True,
parse_defline=False)
names = set([])
for s in a:
names.add(s.name)
tier1_cnt = 0
tier2_cnt = 0
seqs = []
for name in names:
try:
seq = db100_map[name]
info = ''
if name in tier1_list:
tier1_cnt += 1
#info = seq.info + ' ' + tier1_annots[name]
elif name in tier2:
tier2_cnt += 1
#info = seq.info + ' ' + tier2_annots[name]
seqs.append(
sequence.Sequence(seq.sequence, seq.alphabet, seq.name,
info))
except:
pass
#print('Did not find', name)
print('Processed', len(seqs), 'for', result_file, ' Tier-1:',
tier1_cnt, ' Tier-2:', tier2_cnt)
output = [ev, red, len(seqs)]
totalSeqCount.append(output)
plotSeqs(totalSeqCount)
print "Q5 sequence has ", len(sequences_q5), " entries"
print "Q6 sequence has ", len(sequences_q6), " entries"
ids_q5 = []
ids_q6 = []
for sequence in sequences_q5:
ids_q5.append(sequence.name)
for sequence in sequences_q6:
ids_q6.append(sequence.name)
common_ids = set(ids_q5).intersection(set(ids_q6))
print len(common_ids), " common matches found"
""" save the common entries into a FASTA file as well as a dictionary
of id:sequence object map
"""
result_sequences = []
result_dictionary = {}
for sequence in sequences_q5:
if sequence.name in common_ids:
result_sequences.append(sequence)
result_dictionary[sequence.name] = sequence
seq.writeFastaFile(ex7_filename, result_sequences)
print "saved results to ", ex7_filename
print('Warning: ', nodename, 'has not got a sequence')
all_children = tree.getDescendantsOf(node, transitive=True)
if all_children == None:
print('Warning: ', nodename, 'has no children')
direct_ancestors = tree.getAncestorsOf(node, transitive=True)
if direct_ancestors == None:
print('Warning: ', nodename, 'has no ancestors')
direct_ancestors = []
relevant = [node.sequence]
for child in all_children:
if not child.sequence:
pass
#print 'Warning: ', nodename, 'has a child', child.label, 'with a sequence which is None'
else:
allgaps = True
for pos in child.sequence:
if pos != '-':
allgaps = False
break
if not allgaps:
relevant.append(child.sequence)
for parent in direct_ancestors:
if not parent.sequence:
print('Warning: ', nodename, 'has an ancestor', parent.label,
'with a sequence which is None')
else:
relevant.append(parent.sequence)
relevant_aln = sequence.Alignment(relevant)
saveConsensus(relevant_aln, countgaps=True, filename=nodename + ".txt")
sequence.writeFastaFile(nodename + ".fa", relevant_aln.seqs)
