def gene_patches(tx, tx_dict, ax, arrow=False):
iso_list = [x for x in tx_dict if tx in x]
if len(iso_list) == 0:
return None
for n, iso in enumerate(iso_list):
start, end, strand, CDS_start, CDS_end, exons, chrom = SP.tx_info(iso, tx_dict)
if arrow is False:
tx_patch = patches.Rectangle((start,0.8-n*0.15),end-start,0.04,edgecolor='0.1',facecolor='0.1')
ax.add_patch(tx_patch)
else:
if strand == '+':
ax.arrow(start, 0.9, end-start-0.02*(end-start), 0, linewidth=2, head_width=0.1,
head_length=0.02*(end-start), fc='k', ec='k')
elif strand == '-':
ax.arrow(end, 0.9, start-end-0.02*(start-end), 0, linewidth=2, head_width=0.1,
head_length=0.02*(end-start), fc='k', ec='k')
if exons is not None:
exon_patches = []
for exon_start, exon_stop in exons:
exon_patches.append(patches.Rectangle((exon_start, 0.775-n*0.15), exon_stop-exon_start, 0.10,
edgecolor='0.1',facecolor='0.1'))
for patch in exon_patches:
ax.add_patch(patch)
else:
CDS_patch = patches.Rectangle((CDS_start, 0.75-n*0.15),CDS_end-CDS_start, 0.10, edgecolor='0.1', facecolor='0.1')
ax.add_patch(CDS_patch)
ax.get_yaxis().set_ticks([])
return strand
def gene_patches(tx, tx_dict, ax, arrow=False):
iso_list = [x for x in tx_dict if tx in x]
if len(iso_list) == 0:
return None
for n, iso in enumerate(iso_list):
start, end, strand, CDS_start, CDS_end, exons, chrom = SP.tx_info(
iso, tx_dict)
if arrow is False:
tx_patch = patches.Rectangle((start, 0.8 - n * 0.15),
end - start,
0.04,
edgecolor='0.1',
facecolor='0.1')
ax.add_patch(tx_patch)
else:
if strand == '+':
ax.arrow(start,
0.9,
end - start - 0.02 * (end - start),
0,
linewidth=2,
head_width=0.1,
head_length=0.02 * (end - start),
fc='k',
ec='k')
elif strand == '-':
ax.arrow(end,
0.9,
start - end - 0.02 * (start - end),
0,
linewidth=2,
head_width=0.1,
head_length=0.02 * (end - start),
fc='k',
ec='k')
if exons is not None:
exon_patches = []
for exon_start, exon_stop in exons:
exon_patches.append(
patches.Rectangle((exon_start, 0.775 - n * 0.15),
exon_stop - exon_start,
0.10,
edgecolor='0.1',
facecolor='0.1'))
for patch in exon_patches:
ax.add_patch(patch)
else:
CDS_patch = patches.Rectangle((CDS_start, 0.75 - n * 0.15),
CDS_end - CDS_start,
0.10,
edgecolor='0.1',
facecolor='0.1')
ax.add_patch(CDS_patch)
ax.get_yaxis().set_ticks([])
return strand
def count_reads_in_transcript(bam_files, df, gff3, organism=None):
tx_dict = SP.build_transcript_dict(gff3, organism=organism)
bams = {}
for bam_file in bam_files:
bams[bam_file] = pysam.Samfile(bam_file)
all_reads = {}
for bam, reader in bams.iteritems():
all_reads[bam] = pd.DataFrame(index=df.index, columns=['total','intron'])
for tx in set(df['transcript']):
tx_df = df[df['transcript'] == tx]
if organism == 'pombe':
tx = tx+'.1'
else:
tx = tx+'T0'
start, end, strand, CDS_start, CDS_end, exons, chrom = SP.tx_info(tx, tx_dict)
if organism == 'pombe':
lat_rom = {'chr1':'I','chr2':'II','chr3':'III'}
chrom = lat_rom[chrom]
tx_iter = reader.fetch(chrom, start, end)
intron_ranges = {}
for ix, r in tx_df.iterrows():
if strand == '+':
intron_start = int(r['position'])
intron_end = int(r['position']+r['intron size'])+1
elif strand == '-':
intron_start = int(r['position']-r['intron size'])
intron_end = int(r['position'])+1
intron_ranges[ix] = [range(intron_start,intron_end),0]
reads = 0
for read in tx_iter:
if read.is_reverse and strand == '+':
reads += 1
for ix in intron_ranges:
if read.reference_end in intron_ranges[ix][0]:
intron_ranges[ix][1] += 1
elif not read.is_reverse and strand == '-':
reads += 1
for ix in intron_ranges:
if read.reference_start in intron_ranges[ix][0]:
intron_ranges[ix][1] += 1
for ix in intron_ranges:
try:
all_reads[bam].loc[ix,'total'] = reads/float(end-start)*1000
all_reads[bam].loc[ix,'intron'] = ((intron_ranges[ix][1]/float(tx_df.loc[ix,'intron size'])) /
(reads/float(end-start)))
except ZeroDivisionError:
all_reads[bam].loc[ix,'total'] = np.NaN
all_reads[bam].loc[ix,'intron'] = np.NaN
print ix
return all_reads
def count_reads_in_transcript(bam_files, df, gff3, organism=None):
tx_dict = SP.build_transcript_dict(gff3, organism=organism)
bams = {}
for bam_file in bam_files:
bams[bam_file] = pysam.Samfile(bam_file)
all_reads = {}
for bam, reader in bams.iteritems():
all_reads[bam] = pd.DataFrame(index=df.index,
columns=['total', 'intron'])
for tx in set(df['transcript']):
tx_df = df[df['transcript'] == tx]
if organism == 'pombe':
tx = tx + '.1'
else:
tx = tx + 'T0'
start, end, strand, CDS_start, CDS_end, exons, chrom = SP.tx_info(
tx, tx_dict)
if organism == 'pombe':
lat_rom = {'chr1': 'I', 'chr2': 'II', 'chr3': 'III'}
chrom = lat_rom[chrom]
tx_iter = reader.fetch(chrom, start, end)
intron_ranges = {}
for ix, r in tx_df.iterrows():
if strand == '+':
intron_start = int(r['position'])
intron_end = int(r['position'] + r['intron size']) + 1
elif strand == '-':
intron_start = int(r['position'] - r['intron size'])
intron_end = int(r['position']) + 1
intron_ranges[ix] = [range(intron_start, intron_end), 0]
reads = 0
for read in tx_iter:
if read.is_reverse and strand == '+':
reads += 1
for ix in intron_ranges:
if read.reference_end in intron_ranges[ix][0]:
intron_ranges[ix][1] += 1
elif not read.is_reverse and strand == '-':
reads += 1
for ix in intron_ranges:
if read.reference_start in intron_ranges[ix][0]:
intron_ranges[ix][1] += 1
for ix in intron_ranges:
try:
all_reads[bam].loc[ix,
'total'] = reads / float(end -
start) * 1000
all_reads[bam].loc[ix, 'intron'] = (
(intron_ranges[ix][1] /
float(tx_df.loc[ix, 'intron size'])) /
(reads / float(end - start)))
except ZeroDivisionError:
all_reads[bam].loc[ix, 'total'] = np.NaN
all_reads[bam].loc[ix, 'intron'] = np.NaN
print ix
return all_reads
