def get_gtf(cur, species):
#print "SELECT seqID, start_pos, end_pos, strand, geneID FROM CodingSequences WHERE species = %s" % (species)
cur.execute("SELECT seqID, start_pos, end_pos, strand, geneID FROM CodingSequences WHERE species = %s", (species))
for row in cur.fetchall():
gtf = { 'seqname': row[0],
'source': 'transdecoder',
'feature': 'CDS',
'start': row[1],
'end': row[2],
'strand': row[3],
'frame': '.',
'gene_id': row[4],
'transcript_id': row[4] + '.1'
}
print flatten_GTF(gtf)
def get_gtf(cur, species):
#print "SELECT seqID, start_pos, end_pos, strand, geneID FROM CodingSequences WHERE species = %s" % (species)
cur.execute(
"SELECT seqID, start_pos, end_pos, strand, geneID FROM CodingSequences WHERE species = %s",
(species))
for row in cur.fetchall():
gtf = {
'seqname': row[0],
'source': 'transdecoder',
'feature': 'CDS',
'start': row[1],
'end': row[2],
'strand': row[3],
'frame': '.',
'gene_id': row[4],
'transcript_id': row[4] + '.1'
}
print flatten_GTF(gtf)
def main(argv):
gtf_filename = ''
species = ''
seqfilename = ''
usage = 'Blast2GTF.py -n <species_name> -g <gtf_outfile> -s <seq_file>'
try:
opts, args = getopt.getopt(argv,"hn:g:s:",["name=", "gtf=", "seqfile="])
if not opts:
raise getopt.GetoptError('no opts')
except getopt.GetoptError:
print usage
sys.exit(2)
for opt, arg in opts:
if opt == "-h":
print usage
sys.exit()
elif opt in ("-n", "--name"):
species = arg
elif opt in ("-s", "--seqfile"):
seqfilename = arg
elif opt in ("-g", "--gtf"):
gtf_filename = arg
con = mdb.connect('localhost', 'root', '', 'Selaginella');
with con:
#make a list of names to ensure that there are no duplicates
name_list = {}
#read dictionary of cluster membership (It would be far better to put this info into the db, but this is working, so I won't mess with it
cluster_info = path.join(path.expanduser("~"), "Bioinformatics", "Selaginella", "RefSeq", "SeqClusters.p")
seq_groups = pickle.load( open( cluster_info, "rb" ) )
if gtf_filename:
outfile = open(gtf_filename, 'wb')
gtf_writer = csv.writer(outfile, delimiter='\t', quotechar='', quoting=csv.QUOTE_NONE)
if seqfilename:
seqfile = open(seqfilename, 'wb')
cur = con.cursor()
cur.execute("SELECT a.seqid, b.sequence FROM Species a, Ortholog_groups b WHERE a.seqid = b.seqid AND a.species= %s", (species))
#cur.execute("SELECT seqid, sequence FROM Ortholog_groups b WHERE seqid = 'UNCcomp100582_c0_seq1'")
rows = cur.fetchall()
for (seqid, seq) in rows:
try:
seq_record = SeqRecord(Seq(seq))
except TypeError:
warnings.warn("Can't create seq object from %s for %s" % (seq, seqid))
continue
if gtf_filename: #Write blast info to GTF file
hit_list = {}
hit_order = []
cur.execute("SELECT * FROM BLAST WHERE qseqid=%s", (seqid,))
for (id, qseqid, qlen, sacc, slen, pident, length, mismatch, gapopen, qstart, qend, qframe, sstart, send, sframe, evalue, bitscore, strand, hitnum) in cur.fetchall():
feature = {
'source': '1kp',
'feature': 'blast_hit',
'frame': '.',
'seqname': qseqid,
'score': float(bitscore),
'start': int(qstart),
'end': int(qend)
}
#Add strand information and reverse coordinates if on negative strand
if strand == '1':
feature['strand'] = '+'
elif strand == '0':
feature['strand'] = '-'
else:
sys.exit("Strand %s not recognized" % strand)
#In cases where multiple hits to the same subject, combine by maximizing coordinates
if sacc in hit_list:
if hit_list[sacc]['start'] > feature['start']:
hit_list[sacc]['start'] = feature['start']
if hit_list[sacc]['end'] < feature['end']:
hit_list[sacc]['end'] = feature['end']
else:
hit_order.append(sacc)
hit_list[sacc] = feature
#Scan through hits in order and write CDS features for non-overlapping ones
for i in range(len(hit_order)):
overlap = 0
sacc = hit_order[i]
feature = hit_list[sacc]
for j in range(i):
overlap = max(overlap, hit_overlap(feature, hit_list[hit_order[j]]))
if not overlap:
if sacc in seq_groups:
name = "%s_c%s_0" % (qseqid[0:qseqid.find('comp')], seq_groups[sacc].split("_")[1])
else:
name = sacc + "_0"
name_num = 1
while name in name_list:
name = name.split("_")[0:-1] + [str(name_num)]
name = "_".join(name)
name_num = name_num + 1
name_list[name] = 1
feature['gene_id'] = name
feature['transcript_id'] = feature['gene_id'] + '.1'
feature['feature'] = 'CDS'
feature['score'] = '.'
if feature['strand'] == '+':
(feature['start'], feature['end']) = get_orf_coords(seq_record, feature['start'], feature['end'])
feature['frame'] = feature['start'] % 3 + 1
#print "feature start, feature end = %s, %s" % ( feature['start'], feature['end'] )
note = orf_integrity(Seq(seq[feature['start']-1:feature['end']]))
if note: feature['note'] = note
else:
(orf_start, orf_end) = get_orf_coords(seq_record.reverse_complement(), len(seq_record) - qend + 1, len(seq_record) - qstart + 1)
(feature['start'], feature['end']) = (len(seq_record) - orf_end + 1, len(seq_record) - orf_start + 1)
feature['frame'] = ( len(seq_record) - feature['end'] + 1 ) % 3 + 1
orf_rev = Seq(seq[feature['start']-1:feature['end']] )
note = orf_integrity(orf_rev.reverse_complement())
if note: feature['note'] = note
gtf_writer.writerow(flatten_GTF(feature))
if seqfilename:
seqfile.write(">%s\n%s\n" % (seqid, seq))