def make_pair_maps(pair_file, fmt, qbed, sbed):
"""make dicts of q => s and s => q"""
qmap_tuple = []
for pair in get_pair(pair_file,fmt, qbed, sbed):
if pair is None: break
(sname, qname) = pair
qmap_tuple.append((qname,sname))
qmap_tuple.append((sname,qname))
return qmap_tuple
def main(qbed, sbed, pairs_file, qpad, spad, unmasked_fasta, pair_fmt,blast_path, mask='F', ncpu=8):
"""main runner for finding cnss"""
pool = Pool(ncpu)
bl2seq = "%s " % blast_path + \
"-p blastn -D 1 -E 2 -q -2 -r 1 -G 5 -W 7 -F %s " % mask + \
" -e %(e_value).2f -i %(qfasta)s -j %(sfasta)s \
-I %(qstart)d,%(qstop)d -J %(sstart)d,%(sstop)d | grep -v '#' \
| grep -v 'WARNING' | grep -v 'ERROR' "
fcnss = sys.stdout
print >> fcnss,
"#qseqid,qaccn,sseqid,saccn,[qstart,qend,sstart,send,bitscore...]"
qfastas = get_masked_fastas(qbed)
sfastas = get_masked_fastas(sbed) if qbed.filename != sbed.filename else qfastas
pairs = [True]
_get_pair_gen = get_pair(pairs_file, pair_fmt, qbed, sbed)
# need this for parallization stuff.
def get_pair_gen():
try: return _get_pair_gen.next()
except StopIteration: return None
while any(pairs):
pairs = [get_pair_gen() for i in range(ncpu)]
# this helps in parallelizing.
spad_map = [spad] * len(pairs)
qpad_map = [qpad] * len(pairs)
sfastas_map = [sfastas] * len(pairs)
qfastas_map = [qfastas] * len(pairs)
bl2seq_map = [bl2seq] * len(pairs)
####################################
cmds = [c for c in map(get_cmd, [l for l in pairs if
l],bl2seq_map,qfastas_map,sfastas_map,qpad_map,spad_map) if c]
results = (r for r in pool.map(commands.getoutput, [c[0] for c in cmds]))
for res, (cmd, qfeat, sfeat) in zip(results, cmds):
if not res.strip(): continue
print >>sys.stderr, "%s %s" % (qfeat["accn"], sfeat['accn']),
orient = qfeat['strand'] == sfeat['strand'] and 1 or -1
cnss = parse_blast(res, orient, qfeat, sfeat, qbed, sbed, qpad, spad, unmasked_fasta)
print >>sys.stderr, "(%i)" % len(cnss)
if len(cnss) == 0: continue
qname, sname = qfeat['accn'], sfeat['accn']
print >> fcnss, "%s,%s,%s,%s,%s" % (qfeat['seqid'], qname, sfeat['seqid'], sname,
",".join(map(lambda l: ",".join(map(str,l)),cnss)))
return None
def make_pair_maps(pair_file, fmt, qbed, sbed):
"""
make dicts of q => s and s => q
"""
qmap = collections.defaultdict(list) # key is query, value is a list of subject hits
smap = collections.defaultdict(list)
print >>sys.stderr, "pair file:", pair_file
for pair in get_pair(pair_file, fmt, qbed, sbed):
if pair is None: break
(qname, sname) = pair
qmap[qname].append(sname)
smap[sname].append(qname)
return qmap, smap
def make_pair_maps(pair_file, fmt, qbed, sbed):
"""
make dicts of q => s and s => q
"""
qmap = collections.defaultdict(
list) # key is query, value is a list of subject hits
smap = collections.defaultdict(list)
print >> sys.stderr, "pair file:", pair_file
for pair in get_pair(pair_file, fmt, qbed, sbed):
if pair is None: break
(qname, sname) = pair
qmap[qname].append(sname)
smap[sname].append(qname)
return qmap, smap
def get_homeolog(qfeat,pairsfile, sbed, qbed):
for region, sregion in get_pair(pairsfile, 'pck', sbed, qbed):
if region['sfeat'] == qfeat[3]:
return region['ORG2_qfeat']
def main(qbed,sbed,missed_pairs, ncpu):
"""run tblastx on missed pairs..."""
#print >>sys.stderr,ncpu
ncpu = int(ncpu)
pool = Pool(ncpu)
pairs_file = get_pairs_file(missed_pairs)
print >>sys.stdout, "#hit,ref_gene,blastn_introns,blastx_hits, blastx_gene_hits, blastx_frame, blastn_gaps, blastx_gaps,orf_perdiction,orf_blastx,frame_shift"
blastn = "/Users/gturco/blast-2.2.25/bin/bl2seq -p blastn -G 5 -E 2 -W 7 -q -2 -e 0.001 -D 1 -i {0} -j {1} -I {2},{3} -J {4},{5} | grep -v '#' | grep -v 'WARNING' | grep -v 'ERROR' "
qfastas = split_fastas(qbed)#MASK CODING
sfastas = get_mask_non_cds(sbed) #mask noncoding
pairs = [True]
_get_pair_gen = get_pair(pairs_file,"pair", qbed,sbed)
def get_pair_gen():
try: return _get_pair_gen.next()
except StopIteration: return None
while any(pairs):
pairs = [get_pair_gen() for i in range(ncpu)]
def get_blastn_cmd(pair):
"""creates the dictionary values used to fill in blast cmd"""
if pair is None: return None
hit, gene = pair
hstart, hstop = abs(3000 - hit['start']), (3000 + hit['end'])
# double check fasta to make sure i dont need to add or remove one
gstart,gstop = gene['start'],gene['end']
# checks the entire gene...
query_file = qfastas[hit['seqid']]
subject_file = sfastas[gene['seqid']]
blastn_cmd = blastn.format(query_file, subject_file, hstart, hstop, gstart, gstop)
#print >> sys.stderr,'{0},{1},{2}'.format(hit['accn'],gene['accn'],cmd)
return blastn_cmd,hit, gene
cmds = [c for c in map(get_blastn_cmd, [l for l in pairs if l]) if c]
#print >>sys.stderr, "results: {0}".format(cmds[0][0])
results = (r for r in pool.map(commands.getoutput,[c[0] for c in cmds]))
for res, (cmd, hit, gene) in zip(results,cmds):
print >>sys.stderr, "CMD: {0},{1}".format(gene['accn'],hit['accn'])
d,no_res = group_cds(res, gene)
gap_list =[]
intron_list = []
hit['locs'] = []
if no_res == True: continue
for group_key in d.keys():
exon_hits = d[group_key]
non_crossing = remove_crossing_hits(exon_hits,hit,gene)
if len(non_crossing) > 1:
gaps,hstart,hend =bites(non_crossing)
gap_list.append(sum(gaps))
elif len(non_crossing) == 1:
# print >>sys.stderr, non_crossing
[(hstart,hend,sstart,send,evalue)] = non_crossing
if len(non_crossing) >= 1:
intron_list.append(group_key[0])
hit['locs'].append((hstart,hend))
hit['locs'].sort()
#print >>sys.stderr, "hit_loc : {0}".format(hit['locs'])
if len(hit['locs']) < 1: continue
orf_prediction = find_orf(qbed,hit)
introns = "{0}/{1}".format(len(intron_list),len(gene['locs']))
gap_totaln = sum(gap_list)
# new hit locs made from blastn res
hit_percent, gene_percent, frame_percent,frame_shift, best_frame, gap_total,orf_start= protein_parse(hit,gene,sbed,qbed)
orf_start = abs(min(hit['locs'][0]) + int(orf_start))
w ="{0},{1},{2},{3},{4},{5},{6},{7},{8},{9},{10}".format(hit['accn'],gene['accn'],introns,hit_percent,gene_percent, frame_percent,gap_totaln,gap_total,orf_prediction,orf_start,frame_shift)
print >>sys.stdout, w