def annotate(df, cols):
""" get longest transcript for gene column or regions """
def get_regions(row, cols):
prefix = '' if row[cols[0]].startswith('chr') else 'chr'
return "{}{}:{}-{}".format(prefix, row[cols[0]], row[cols[1]],
row[cols[2]])
llama = LlamaEnsembl()
ucsc = UCSCapi()
dd = {'query': [], 'transcripts': []}
if len(cols) == 1:
genes = df[cols[0]].values
for gene in genes:
chrom, start, end = llama.get_gene_pos(gene)
res = ucsc.query("chr{}:{}-{}".format(chrom, start, end))
dd['query'].append(gene)
dd['transcripts'].append(res.longest(gene)['transcript'])
mergecol = cols[0]
elif len(cols) == 3:
dd['genes'] = []
df['region'] = df.apply(get_regions, axis=1, args=[cols])
regions = df['region'].values
for region in regions:
res = ucsc.query(region)
for gene in res.genes():
dd['genes'].append(gene)
dd['query'].append(region)
dd['transcripts'].append(res.longest(gene)['transcript'])
mergecol = 'region'
ndf = df.merge(pd.DataFrame(dd),
left_on=mergecol,
right_on='query',
how='outer')
return ndf[~ndf.duplicated()]
def test_gene_transcripts():
llama = LlamaEnsembl()
ucsc = UCSCapi()
gene = 'BRCA1'
chrom, start, end = llama.get_gene_pos(gene)
res = ucsc.query("chr{}:{}-{}".format(chrom, start, end))
transcript = res.longest(gene)
print(transcript['transcript'])
def annotate(self):
"""
Annotate data frame
:param df: pandas dataframe
:return: merged dataframe
"""
llama = LlamaEnsembl()
new_df = llama.annotate_dataframe(self.df)
return self.df.join(new_df)
def test_annotation_hg38():
llama = LlamaEnsembl(genome='hg38')
df = pd.DataFrame({
'CHROM': ['chr1', 'chr17', '7', '7'],
'START': [153358330, 43092618, 116771890, 116773071],
'END': [153358350, 43092648, 116771920, 116773072],
'gene_exp': ['S100A9', 'BRCA1', 'MET', 'MET'],
'exon_exp': ['2', '10', '14', ''],
'strand_exp': ['+', '-', '+', '+']
})
ndf = llama.annotate_dataframe(df)
mdf = pd.concat([df, ndf], axis=1)
print(mdf)
assert (mdf[mdf['genes'] != mdf['gene_exp']].shape[0] == 0)
assert (mdf[mdf['exons'] != mdf['exon_exp']].shape[0] == 0)
def test_annotation():
llama = LlamaEnsembl()
df = pd.DataFrame({
'CHROM': ['chr1', 'chr17', '7', '7'],
'START': [153330766, 41244573, 116412043, 116412044],
'END': [153330797, 41245953, 116412043, 116412044],
'gene_exp': ['S100A9', 'BRCA1', 'MET', 'MET'],
'exon_exp': ['2', '10', '14', ''],
'strand_exp': ['+', '-', '+', '+']
})
ndf = llama.annotate_dataframe(df)
mdf = pd.concat([df, ndf], axis=1)
print(mdf)
assert (mdf[mdf['genes'] != mdf['gene_exp']].shape[0] == 0)
assert (mdf[mdf['exons'] != mdf['exon_exp']].shape[0] == 0)
def test_rsid():
llama = LlamaEnsembl()
rsids = """rs1517114
rs4646
rs55886062
rs3918290
rs67376798
rs75017182
rs115232898
rs1801158
rs11615
rs1800566
rs7779029
rs151264360
rs4148323
rs8175347"""
rsidstr = [r.lstrip() for r in rsids.split('\n')]
df = llama.annotate_variants(rsidstr, extra_cols=['MAF', 'ambiguity'])
print(df)
def test_cds_convert():
llama = LlamaEnsembl()
result = llama.get_cds_region('NM_015506.2', 'c.1A>G')
result2 = llama.get_cds_region('NM_015506.2', 'c.445_446del')
print(result)
print(result2)