def checkGenes(chromosomes, exon_records, chromosome_load_fxn, cdnas, id_fxn):
total_genes = 0
total_errors = 0
# Reconstruct genes
for chr in chromosomes:
print "# Chromosome", chr
# Load this chromosome's data
chrseq = chromosome_load_fxn(chr)
n_genes = 0
n_errors = 0
n_wrong_length = 0
n_bad_translation = 0
n_different_translation = 0
n_no_comparison = 0
for (pepid, recdict) in exon_records.items():
recs = recdict.values()
if recs[0].chromosome == chr:
recs.sort(key=lambda e: e.exon_start)
(seq, sentinel_rec) = buildCodingSequence(recs, chrseq, 0)
# Reverse-complement the sequence if it's on the negative strand
if recs[0].strand == '-1':
#print "reversing strand"
seq = translate.reverseComplement(seq)
n_genes += 1
try:
orf = id_fxn(sentinel_rec)
cdna = cdnas[orf]
#print orf
#print seq
#print cdna
if len(seq) == len(cdna):
if seq != cdna:
n_errors += 1
elif len(seq) - 3 == len(cdna):
if seq[0:-3] != cdna:
n_errors += 1
else:
n_errors += 1
n_wrong_length += 1
except KeyError:
n_no_comparison += 1
continue
print "# Processed %d genes with %d errors" % (n_genes, n_errors)
print "# %d length errors" % (n_wrong_length, )
print "# %d no comparison errors" % (n_no_comparison, )
print "# %d bad translation errors" % (n_bad_translation, )
print "# %d different translation errors" % (
n_different_translation, )
total_errors += n_errors
total_genes += n_genes
print "# Processed %d genes with %d errors total" % (total_genes,
total_errors)
def checkGenes(chromosomes, exon_records, chromosome_load_fxn, cdnas, id_fxn): total_genes = 0 total_errors = 0 # Reconstruct genes for chr in chromosomes: print "# Chromosome", chr # Load this chromosome's data chrseq = chromosome_load_fxn(chr) n_genes = 0 n_errors = 0 n_wrong_length = 0 n_bad_translation = 0 n_different_translation = 0 n_no_comparison = 0 for (pepid, recdict) in exon_records.items(): recs = recdict.values() if recs[0].chromosome == chr: recs.sort( key = lambda e: e.exon_start) (seq, sentinel_rec) = buildCodingSequence(recs, chrseq, 0) # Reverse-complement the sequence if it's on the negative strand if recs[0].strand == '-1': #print "reversing strand" seq = translate.reverseComplement(seq) n_genes += 1 try: orf = id_fxn(sentinel_rec) cdna = cdnas[orf] #print orf #print seq #print cdna if len(seq) == len(cdna): if seq != cdna: n_errors += 1 elif len(seq)-3 == len(cdna): if seq[0:-3] != cdna: n_errors += 1 else: n_errors += 1 n_wrong_length += 1 except KeyError: n_no_comparison += 1 continue print "# Processed %d genes with %d errors" % (n_genes, n_errors) print "# %d length errors" % (n_wrong_length,) print "# %d no comparison errors" % (n_no_comparison,) print "# %d bad translation errors" % (n_bad_translation,) print "# %d different translation errors" % (n_different_translation,) total_errors += n_errors total_genes += n_genes print "# Processed %d genes with %d errors total" % (total_genes, total_errors)
def getSequence(self, raw_seq, n_bases_upstream=0, n_bases_downstream=0): # Sort exon records self.exons.sort(key=lambda x: x.coding_start) # Figure out which direction "upstream" is if self.strand == '-': # Swap -- downstream is really upstream on the negative strand (n_bases_upstream, n_bases_downstream) = (n_bases_downstream, n_bases_upstream) seq = '' if len(self.exons) == 1: seq += self.exons[0].pullCodingSequence(raw_seq, n_bases_upstream, n_bases_downstream) else: for (i,x) in enumerate(self.exons): if i==0: seq += x.pullCodingSequence(raw_seq, n_bases_upstream, 0) elif i==(len(self.exons)-1): seq += x.pullCodingSequence(raw_seq, 0, n_bases_downstream) else: seq += x.pullCodingSequence(raw_seq, 0, 0) if self.strand == '-': seq = translate.reverseComplement(seq) return seq
def getSequence(self, raw_seq, n_bases_upstream=0, n_bases_downstream=0):
# Sort exon records
self.exons.sort(key=lambda x: x.coding_start)
# Figure out which direction "upstream" is
if self.strand == '-':
# Swap -- downstream is really upstream on the negative strand
(n_bases_upstream, n_bases_downstream) = (n_bases_downstream,
n_bases_upstream)
seq = ''
if len(self.exons) == 1:
seq += self.exons[0].pullCodingSequence(raw_seq, n_bases_upstream,
n_bases_downstream)
else:
for (i, x) in enumerate(self.exons):
if i == 0:
seq += x.pullCodingSequence(raw_seq, n_bases_upstream, 0)
elif i == (len(self.exons) - 1):
seq += x.pullCodingSequence(raw_seq, 0, n_bases_downstream)
else:
seq += x.pullCodingSequence(raw_seq, 0, 0)
if self.strand == '-':
seq = translate.reverseComplement(seq)
return seq
def getStrandAlts(self, strand): res = self.alt if strand == '-': res = [translate.reverseComplement(a) for a in self.alt] return res
def getStrandRef(self, strand): res = self.ref if strand == '-': res = translate.reverseComplement(self.ref) return res
def getStrandAlts(self, strand):
res = self.alt
if strand == '-':
res = [translate.reverseComplement(a) for a in self.alt]
return res
def getStrandRef(self, strand):
res = self.ref
if strand == '-':
res = translate.reverseComplement(self.ref)
return res
def test002(self):
"""reverse complement"""
s = "ATGCatgc"
self.assertTrue(translate.reverseComplement(s) == "gcatGCAT")
self.assertTrue(translate.reverseComplement(translate.reverseComplement(s)) == s)
def writeGenes(chromosomes, exon_records, chromosome_load_fxn, id_fxn, exclude_nt_from_boundary, outfile):
total_genes = 0
total_errors = 0
# Reconstruct genes
for chr in chromosomes:
print "# Chromosome", chr
# Load this chromosome's data
chrseq = chromosome_load_fxn(chr)
n_genes = 0
n_errors = 0
n_wrong_length = 0
n_bad_translation = 0
n_different_translation = 0
n_no_comparison = 0
for (pepid, recdict) in exon_records.items():
recs = recdict.values()
if recs[0].chromosome == chr:
strand_sign = int(recs[0].strand)
recs.sort( key = lambda e: e.exon_start)
(seq, sentinel_rec) = buildCodingSequence(recs, chrseq, exclude_nt_from_boundary)
rseq = seq
# Reverse-complement the sequence if it's on the negative strand
if recs[0].strand == '-1':
#print "reversing strand"
seq = translate.reverseComplement(seq)
n_genes += 1
if len(seq) % 3 != 0:
n_errors += 1
n_wrong_length += 1
continue
if exclude_nt_from_boundary == 0:
#if sentinel_rec.peptide_ID == 'ENSDARP00000076309':
# print seq
# print rseq
# print translate.translateRaw(seq)
# print translate.translateRaw(rseq)
# sys.exit()
try:
prot = translate.translate(seq)
if False: #not prot:
print "****"
print id_fxn(sentinel_rec)
print seq
print rseq
for rec in recs:
print rec
#print "^%s^\t%s" % (recs[0].strand, seq[0:])
n_errors += 1
n_bad_translation += 1
continue
except translate.BioUtilsError:
continue
# Translation is good... write it.
line = ">%s\n%s\n" % (id_fxn(sentinel_rec), seq)
outfile.write(line)
#outfile.write(">%s\n%s\n" % (id_fxn(sentinel_rec), seq))
print "# Processed %d genes with %d errors" % (n_genes, n_errors)
print "# %d length errors" % (n_wrong_length,)
print "# %d bad translation errors" % (n_bad_translation,)
total_errors += n_errors
total_genes += n_genes
print "# Wrote %d genes with %d errors total" % (total_genes, total_errors)
def writeStats(chromosomes, peptide_records, bounds, chromosome_load_fxn, outfile):
header = "gene\tpep\ttrans\tchr\tlen.intron\tn.exons\tfrac.coding\tgc.intron\tgcind.intron\tgt.intron\tlen.coding\tfrac.%s\n" % \
('\tfrac.'.join([str(b) for b in bounds],))
outfile.write(header)
print header,
total_genes = 0
total_errors = 0
# Reconstruct genes
for chr in chromosomes:
print "# Chromosome", chr
# Load this chromosome's data
chrseq = chromosome_load_fxn(chr)
n_genes = 0
n_errors = 0
n_wrong_length = 0
n_bad_translation = 0
n_different_translation = 0
for (pepid, recdict) in peptide_records.items():
recs = recdict.values()
if recs[0].chromosome == chr:
recs.sort( key = lambda e: e.exon_start)
(seq, sentinel_rec) = buildCodingSequence(recs, chrseq, 0)
(intron_seq, sentinel_rec_intron) = buildIntronSequence(recs, chrseq)
# Reverse-complement the sequence if it's on the negative strand
if recs[0].strand == '-1':
#print "reversing strand"
seq = translate.reverseComplement(seq)
intron_seq = translate.reverseComplement(intron_seq)
n_genes += 1
# Write out statistics
if False:
print 'g', seq
print 'i', intron_seq
if len(intron_seq)>0:
gc_intron = '%1.4f' % cai.getGC(intron_seq)
intron_length = len(intron_seq)
frac_coding = '%1.4f' % (len(seq)/(len(seq)+float(intron_length)),)
gcind_intron = '%1.4f' % cai.getDinucleotideIndex(intron_seq, 'GC')
gt_intron = '%1.4f' % cai.getContent(intron_seq, 'GT')
else:
gc_intron = 'NA'
intron_length = 0
gcind_intron = 'NA'
if len(seq)>0:
frac_coding = '1.0'
else:
frac_coding = 'NA'
num_coding_exons = len([xr for xr in recs if xr.coding_end > xr.coding_start])
# Test to ensure agreement between codingOutsideBoundary and buildCodingSequence
if False:
(coding_outside, total_coding) = codingOutsideBoundary(recs, 0)
print len(seq), total_coding, coding_outside
assert total_coding == coding_outside
assert coding_outside == len(seq)
fracs_inside = []
for bound in bounds:
(coding_outside, total_coding) = codingOutsideBoundary(recs, bound)
if total_coding > 0:
frac_inside_bound = "%1.4f" % (1-float(coding_outside)/total_coding,)
else:
frac_inside_bound = "NA"
fracs_inside.append(frac_inside_bound)
line = "%s\t%s\t%s\t%s\t%d\t%d\t%s\t%s\t%s\t%s\t%d\t%s\n" % \
(sentinel_rec.gene_ID, sentinel_rec.peptide_ID, sentinel_rec.transcript_ID, sentinel_rec.chromosome, intron_length, num_coding_exons, frac_coding, gc_intron, gcind_intron, gt_intron, len(seq), '\t'.join(fracs_inside))
outfile.write(line)
#print line,
outfile.flush()
print "# Processed %d genes with %d errors" % (n_genes, n_errors)
print "# %d length errors" % (n_wrong_length,)
print "# %d bad translation errors" % (n_bad_translation,)
print "# %d different translation errors" % (n_different_translation,)
total_errors += n_errors
total_genes += n_genes
chrseqlist = None
chrseq = None
print "# Processed %d genes with %d errors total" % (total_genes, total_errors)
outfile.close()
def test002(self):
"""reverse complement"""
s = 'ATGCatgc'
self.assertTrue(translate.reverseComplement(s) == 'gcatGCAT')
self.assertTrue(
translate.reverseComplement(translate.reverseComplement(s)) == s)
def writeGenes(chromosomes, exon_records, chromosome_load_fxn, id_fxn,
exclude_nt_from_boundary, outfile):
total_genes = 0
total_errors = 0
# Reconstruct genes
for chr in chromosomes:
print "# Chromosome", chr
# Load this chromosome's data
chrseq = chromosome_load_fxn(chr)
n_genes = 0
n_errors = 0
n_wrong_length = 0
n_bad_translation = 0
n_different_translation = 0
n_no_comparison = 0
for (pepid, recdict) in exon_records.items():
recs = recdict.values()
if recs[0].chromosome == chr:
strand_sign = int(recs[0].strand)
recs.sort(key=lambda e: e.exon_start)
(seq,
sentinel_rec) = buildCodingSequence(recs, chrseq,
exclude_nt_from_boundary)
rseq = seq
# Reverse-complement the sequence if it's on the negative strand
if recs[0].strand == '-1':
#print "reversing strand"
seq = translate.reverseComplement(seq)
n_genes += 1
if len(seq) % 3 != 0:
n_errors += 1
n_wrong_length += 1
continue
if exclude_nt_from_boundary == 0:
#if sentinel_rec.peptide_ID == 'ENSDARP00000076309':
# print seq
# print rseq
# print translate.translateRaw(seq)
# print translate.translateRaw(rseq)
# sys.exit()
try:
prot = translate.translate(seq)
if False: #not prot:
print "****"
print id_fxn(sentinel_rec)
print seq
print rseq
for rec in recs:
print rec
#print "^%s^\t%s" % (recs[0].strand, seq[0:])
n_errors += 1
n_bad_translation += 1
continue
except translate.BioUtilsError:
continue
# Translation is good... write it.
line = ">%s\n%s\n" % (id_fxn(sentinel_rec), seq)
outfile.write(line)
#outfile.write(">%s\n%s\n" % (id_fxn(sentinel_rec), seq))
print "# Processed %d genes with %d errors" % (n_genes, n_errors)
print "# %d length errors" % (n_wrong_length, )
print "# %d bad translation errors" % (n_bad_translation, )
total_errors += n_errors
total_genes += n_genes
print "# Wrote %d genes with %d errors total" % (total_genes, total_errors)
def writeStats(chromosomes, peptide_records, bounds, chromosome_load_fxn,
outfile):
header = "gene\tpep\ttrans\tchr\tlen.intron\tn.exons\tfrac.coding\tgc.intron\tgcind.intron\tgt.intron\tlen.coding\tfrac.%s\n" % \
('\tfrac.'.join([str(b) for b in bounds],))
outfile.write(header)
print header,
total_genes = 0
total_errors = 0
# Reconstruct genes
for chr in chromosomes:
print "# Chromosome", chr
# Load this chromosome's data
chrseq = chromosome_load_fxn(chr)
n_genes = 0
n_errors = 0
n_wrong_length = 0
n_bad_translation = 0
n_different_translation = 0
for (pepid, recdict) in peptide_records.items():
recs = recdict.values()
if recs[0].chromosome == chr:
recs.sort(key=lambda e: e.exon_start)
(seq, sentinel_rec) = buildCodingSequence(recs, chrseq, 0)
(intron_seq,
sentinel_rec_intron) = buildIntronSequence(recs, chrseq)
# Reverse-complement the sequence if it's on the negative strand
if recs[0].strand == '-1':
#print "reversing strand"
seq = translate.reverseComplement(seq)
intron_seq = translate.reverseComplement(intron_seq)
n_genes += 1
# Write out statistics
if False:
print 'g', seq
print 'i', intron_seq
if len(intron_seq) > 0:
gc_intron = '%1.4f' % cai.getGC(intron_seq)
intron_length = len(intron_seq)
frac_coding = '%1.4f' % (
len(seq) / (len(seq) + float(intron_length)), )
gcind_intron = '%1.4f' % cai.getDinucleotideIndex(
intron_seq, 'GC')
gt_intron = '%1.4f' % cai.getContent(intron_seq, 'GT')
else:
gc_intron = 'NA'
intron_length = 0
gcind_intron = 'NA'
if len(seq) > 0:
frac_coding = '1.0'
else:
frac_coding = 'NA'
num_coding_exons = len(
[xr for xr in recs if xr.coding_end > xr.coding_start])
# Test to ensure agreement between codingOutsideBoundary and buildCodingSequence
if False:
(coding_outside,
total_coding) = codingOutsideBoundary(recs, 0)
print len(seq), total_coding, coding_outside
assert total_coding == coding_outside
assert coding_outside == len(seq)
fracs_inside = []
for bound in bounds:
(coding_outside,
total_coding) = codingOutsideBoundary(recs, bound)
if total_coding > 0:
frac_inside_bound = "%1.4f" % (
1 - float(coding_outside) / total_coding, )
else:
frac_inside_bound = "NA"
fracs_inside.append(frac_inside_bound)
line = "%s\t%s\t%s\t%s\t%d\t%d\t%s\t%s\t%s\t%s\t%d\t%s\n" % \
(sentinel_rec.gene_ID, sentinel_rec.peptide_ID, sentinel_rec.transcript_ID, sentinel_rec.chromosome, intron_length, num_coding_exons, frac_coding, gc_intron, gcind_intron, gt_intron, len(seq), '\t'.join(fracs_inside))
outfile.write(line)
#print line,
outfile.flush()
print "# Processed %d genes with %d errors" % (n_genes, n_errors)
print "# %d length errors" % (n_wrong_length, )
print "# %d bad translation errors" % (n_bad_translation, )
print "# %d different translation errors" % (
n_different_translation, )
total_errors += n_errors
total_genes += n_genes
chrseqlist = None
chrseq = None
print "# Processed %d genes with %d errors total" % (total_genes,
total_errors)
outfile.close()