def output_exons_as_bed(self):
"""
Output the table's exons as BED.
Only implemented for ensGene.txt; probably not
necessary to work out for other tables since this is
only used for aggregate statistics.
"""
if self.source != "ensGene":
return
output_filename = os.path.join(self.exons_dir,
"%s.exons.bed" %(self.source))
print "Outputting exons..."
if os.path.isfile(output_filename):
print " - Found %s. Skipping..." %(output_filename)
return output_filename
exons_file = open(output_filename, "w")
for idx, series in self.raw_table.iterrows():
gene_info = series.to_dict()
gene_id = gene_info["name2"]
chrom = gene_info["chrom"]
strand = gene_info["strand"]
exonStarts = gene_info["exonStarts"]
exonEnds = gene_info["exonEnds"]
# Keep 0-based start of ensGene table since
# this will be outputted as a BED
exon_starts = (int(start) for start in exonStarts.rstrip(",").split(","))
exon_ends = (int(end) for end in exonEnds.rstrip(",").split(","))
exon_coords = zip(exon_starts, exon_ends)
# Output as BED: encode gene ID
bedtools_utils.output_intervals_as_bed(exons_file,
chrom, exon_coords, strand,
name=gene_id)
exons_file.close()
return output_filename
def output_introns(self, min_intron_size=50):
"""
Given the merged exons, compute the intronic coordinates.
Only implemented for ensGene.txt; probably not
necessary to work out for other tables since this is
only used for aggregate statistics.
Exclude intronic content that is less than 'min_intron_size'.
"""
if self.source != "ensGene":
return
output_filename = os.path.join(self.introns_dir,
"ensGene.introns.bed")
print "Outputting introns..."
if os.path.isfile(output_filename):
print " - Found %s. Skipping..." %(output_filename)
return
print " - Output file: %s" %(output_filename)
introns_file = open(output_filename, "w")
# Load ensGene exons
merged_exons_by_gene = self.load_merged_exons_by_gene()
for gene_id, merged_exons in merged_exons_by_gene.iteritems():
chrom = merged_exons[0]["chrom"]
strand = merged_exons[0]["strand"]
# For each gene, get its list of introns and serialize them
exon_coords = [(int(exon["start"]), int(exon["end"])) \
for exon in merged_exons]
intron_coords = []
for first_exon, second_exon in zip(exon_coords, exon_coords[1::1]):
# Intron start coordinate is the coordinate right after
# the end of the first exon, intron end coordinate is the
# coordinate just before the beginning of the second exon
intron_start = first_exon[1] + 1
intron_end = second_exon[0] - 1
if intron_start >= intron_end:
continue
# Filter on intron size (in 0-based coordinates)
intron_size = intron_end - intron_start
if intron_size < min_intron_size:
continue
intron_coords.append((intron_start, intron_end))
bedtools_utils.output_intervals_as_bed(introns_file,
chrom,
intron_coords,
strand,
name=gene_id)
introns_file.close()