def fasta_stats(fastafile, display=False):
# get basic stats on the fasta file:
genelens = sorted( [ len(seq[1]) for seq in internal.parsefasta(fastafile)] , reverse=True)
totallen = sum(genelens)
print "Number of transcripts: %d" % len(genelens)
print "Total length of transcripts: %d" % totallen
print "longest transcript: %d" % genelens[0]
if display:
plt.hist(genelens)
plt.title("Histogram of gene lengths (bp)")
plt.show()
prevl = 0
lensum = 0
trancount = 0
for l in genelens:
if lensum + l >= (totallen / 2):
print "N50: %d (%d transcripts)" % (prevl, trancount)
break
lensum += l
prevl = l
trancount += 1
def fasta_stats(fastafile, display=False):
# get basic stats on the fasta file:
genelens = sorted([len(seq[1]) for seq in internal.parsefasta(fastafile)],
reverse=True)
totallen = sum(genelens)
print "Number of transcripts: %d" % len(genelens)
print "Total length of transcripts: %d" % totallen
print "longest transcript: %d" % genelens[0]
if display:
plt.hist(genelens)
plt.title("Histogram of gene lengths (bp)")
plt.show()
prevl = 0
lensum = 0
trancount = 0
for l in genelens:
if lensum + l >= (totallen / 2):
print "N50: %d (%d transcripts)" % (prevl, trancount)
break
lensum += l
prevl = l
trancount += 1
if args.stats and args.fasta:
fasta_stats(args.fasta, args.display_on)
#############################################################################
### extract sequence information from each gene in supplied transdecoder file
transcript_dic = {}
gene_families = {}
seq_families = {}
geneid_idx = {}
gf_idx = {}
full_blast = get_full_blast_idx(args.blast)
verbalise("Y", "Created full blast index with %d entries" % (len(full_blast)))
for defline, seq in internal.parsefasta(args.transdecoder):
# get trinity and transdecoder gene ids:
tdid, trinityid = parse_defline(defline)
# get any blast results:
if tdid in full_blast:
blastline = full_blast[tdid]
else:
blastline = None
# create new transcript instance
newtranscript = Transcript(trinity_id=trinityid,
transdecoder_id=tdid,
blastline=blastline,
seq=seq)
def get_similar_sequences(temp_dir,
buildhmmer=False,
fastafile=None,
specieslist={},
species=None,
genes=[],
dbpaths={},
mincollect=2,
globalthresh=0.2,
localthresh=0.8,
verbalise=lambda *a: None):
# clean gene list type and content:
if not isinstance(genes, list):
genes = [genes]
genes = [g for g in genes if g != '']
# count genes provided:
genelist_num, fasta_num = internal.count_genes(genes, fastafile)
verbalise("Y",
"Genelist size:%d\nFasta size:%d" % (genelist_num, fasta_num))
# if fasta files are provided, create a temp fastafile to search against with hmmer:
if fastafile:
extra_file = os.path.join(temp_dir, "query_fasta")
handle = open(extra_file, 'w')
for defline, seq in internal.parsefasta(fastafile):
handle.write(">%s\n%s\n" % (defline, seq))
handle.close()
extra_file_search = extra_file
else:
extra_file_search = None
if genelist_num + fasta_num > 1:
buildhmmer = True
if buildhmmer:
hmminput = os.path.join(temp_dir, "hmminput.fa")
handle = open(hmminput, 'w')
seqcount = 0
verbalise("B",
"Extracting sequence data from %d peptides" % len(genes))
for defline, seq, species in internal.get_gene_fastas(
genes=genes,
species=None,
fastafile=fastafile,
specieslist=specieslist,
dbpaths=dbpaths):
if seq:
seqcount += 1
fasta_seq = "%s\n%s\n" % (defline, seq)
handle.write(fasta_seq)
handle.close()
if seqcount == 0:
verbalise("R", "No gene sequences were found.")
return {}
# create alignment of input sequences:
mafft_align1 = os.path.join(temp_dir, "mafft_align_input.fa")
mafft_align(hmminput, mafft_align1)
verbalise("B",
"Creating hidden markov model from %d sequences" % seqcount)
# create hmmbuild model of alignment:
hmmmodel = os.path.join(temp_dir, "hmmmodel.fa")
open(hmmmodel, 'a').close()
handle = os.popen(" ".join(
['hmmbuild --informat afa', hmmmodel, mafft_align1]))
handle.close()
homologlist = hmmer_search(None,
specieslist,
query_species=species,
minthresh=localthresh,
temp_dir=temp_dir,
dbpaths=dbpaths,
mincollect=mincollect,
globalthresh=globalthresh,
hmmfile=hmmmodel,
verbalise=verbalise,
extra_file_search=extra_file_search)
os.remove(mafft_align1)
os.remove(hmminput)
else:
verbalise("B", "Extracting sequence from %s" % genes)
if not isinstance(genes, list):
genes = [genes]
# run phmmer on a single input gene/sequence:
seq = ""
for defline, seq, species in internal.get_gene_fastas(
genes=genes,
species=species,
fastafile=fastafile,
specieslist=specieslist,
dbpaths=dbpaths):
fasta_seq = "%s\n%s\n" % (defline, seq)
verbalise("C", fasta_seq)
if not seq:
verbalise("R", "No genes sequences were found.")
return {}
## phmmer all lpep files
homologlist = hmmer_search(fasta_seq,
specieslist,
query_species=species,
minthresh=localthresh,
dbpaths=dbpaths,
temp_dir=temp_dir,
mincollect=mincollect,
globalthresh=globalthresh,
hmmfile=None,
verbalise=verbalise,
extra_file_search=extra_file_search)
return homologlist
if args.stats and args.fasta:
fasta_stats(args.fasta, args.display_on)
#############################################################################
### extract sequence information from each gene in supplied transdecoder file
transcript_dic = {}
gene_families = {}
seq_families = {}
geneid_idx = {}
gf_idx = {}
full_blast = get_full_blast_idx(args.blast)
verbalise("Y",
"Created full blast index with %d entries" % (len(full_blast)))
for defline, seq in internal.parsefasta(args.transdecoder):
# get trinity and transdecoder gene ids:
tdid, trinityid = parse_defline(defline)
# get any blast results:
if tdid in full_blast:
blastline = full_blast[tdid]
else:
blastline = None
# create new transcript instance
newtranscript = Transcript(trinity_id=trinityid,
transdecoder_id=tdid,
blastline=blastline,
seq=seq)
def get_similar_sequences(temp_dir, buildhmmer=False, fastafile=None,
specieslist={}, species=None, genes=[], dbpaths={},
mincollect=2, globalthresh=0.2, localthresh=0.8,
verbalise=lambda *a: None):
# clean gene list type and content:
if not isinstance(genes, list):
genes = [genes]
genes = [ g for g in genes if g != '' ]
# count genes provided:
genelist_num, fasta_num = internal.count_genes(genes, fastafile)
verbalise("Y", "Genelist size:%d\nFasta size:%d" % (genelist_num, fasta_num ))
# if fasta files are provided, create a temp fastafile to search against with hmmer:
if fastafile:
extra_file = os.path.join(temp_dir, "query_fasta")
handle = open(extra_file, 'w')
for defline, seq in internal.parsefasta(fastafile):
handle.write(">%s\n%s\n" % (defline, seq))
handle.close()
extra_file_search = extra_file
else:
extra_file_search = None
if genelist_num + fasta_num > 1:
buildhmmer = True
if buildhmmer:
hmminput = os.path.join(temp_dir, "hmminput.fa")
handle = open(hmminput, 'w')
seqcount = 0
verbalise("B", "Extracting sequence data from %d peptides" % len(genes))
for defline, seq, species in internal.get_gene_fastas(genes=genes,
species=None,
fastafile=fastafile,
specieslist=specieslist,
dbpaths=dbpaths):
if seq:
seqcount += 1
fasta_seq = "%s\n%s\n" % (defline, seq)
handle.write(fasta_seq)
handle.close()
if seqcount == 0:
verbalise("R", "No gene sequences were found.")
return {}
# create alignment of input sequences:
mafft_align1 = os.path.join(temp_dir, "mafft_align_input.fa")
mafft_align(hmminput, mafft_align1)
verbalise("B", "Creating hidden markov model from %d sequences" % seqcount)
# create hmmbuild model of alignment:
hmmmodel = os.path.join(temp_dir, "hmmmodel.fa")
open(hmmmodel, 'a').close()
handle = os.popen(" ".join(['hmmbuild --informat afa', hmmmodel, mafft_align1]))
handle.close()
homologlist = hmmer_search(None,
specieslist,
query_species=species,
minthresh=localthresh,
temp_dir=temp_dir,
dbpaths=dbpaths,
mincollect=mincollect,
globalthresh=globalthresh,
hmmfile=hmmmodel,
verbalise=verbalise,
extra_file_search=extra_file_search)
os.remove(mafft_align1)
os.remove(hmminput)
else:
verbalise("B", "Extracting sequence from %s" % genes)
if not isinstance(genes, list):
genes = [genes]
# run phmmer on a single input gene/sequence:
seq = ""
for defline, seq, species in internal.get_gene_fastas(genes=genes,
species=species,
fastafile=fastafile,
specieslist=specieslist,
dbpaths=dbpaths):
fasta_seq = "%s\n%s\n" % (defline, seq)
verbalise("C", fasta_seq)
if not seq:
verbalise("R", "No genes sequences were found.")
return {}
## phmmer all lpep files
homologlist = hmmer_search(fasta_seq,
specieslist,
query_species=species,
minthresh=localthresh,
dbpaths=dbpaths,
temp_dir=temp_dir,
mincollect=mincollect,
globalthresh=globalthresh,
hmmfile=None,
verbalise=verbalise,
extra_file_search=extra_file_search)
return homologlist
verbalise=verbalise)
verbalise(
"C", "Best hmmer score = %d" % max(v[1] for v in homologlist.values()))
######### Extract identified sequences from LNRP fasta files #########
conv_handle = open(logfile[:-3] + 'name_conversion.txt', 'w')
conv_dic = {}
itercount = 0
previousseq = ""
seqdic = {} # loaded up to remove duplicate sequences
excluded_genes = config.make_a_list(args.exclude_genes)
excluded_species = config.make_a_list(args.exclude_species)
# place any sequences provided in the input into the seqdic
if args.fasta:
for defline, seq in internal.parsefasta(args.fasta):
if sequence_filter(seq, args.maxlength, args.minlength):
continue
else:
seqdic[seq] = defline
for homolog in sorted(homologlist):
# remove excluded genes before bothering to look up their sequence:
searchname = internal.fix_leaky_pipes(homolog)
if searchname in excluded_genes:
continue
if homologlist[homolog][0] in excluded_species:
continue
# extract sequences of remaining genes and add to conversion dictionary
itercount += 1
verbalise=verbalise,
)
verbalise("C", "Best hmmer score = %d" % max(v[1] for v in homologlist.values()))
######### Extract identified sequences from LNRP fasta files #########
conv_handle = open(logfile[:-3] + "name_conversion.txt", "w")
conv_dic = {}
itercount = 0
previousseq = ""
seqdic = {} # loaded up to remove duplicate sequences
excluded_genes = config.make_a_list(args.exclude_genes)
excluded_species = config.make_a_list(args.exclude_species)
# place any sequences provided in the input into the seqdic
if args.fasta:
for defline, seq in internal.parsefasta(args.fasta):
if sequence_filter(seq, args.maxlength, args.minlength):
continue
else:
seqdic[seq] = defline
for homolog in sorted(homologlist):
# remove excluded genes before bothering to look up their sequence:
searchname = internal.fix_leaky_pipes(homolog)
if searchname in excluded_genes:
continue
if homologlist[homolog][0] in excluded_species:
continue
# extract sequences of remaining genes and add to conversion dictionary
itercount += 1
