def kegg_gene_sets(org):
""" Returns gene sets from KEGG pathways.
"""
caching.clear_cache()
kegg_org = kegg.KEGGOrganism(taxonomy.name(org))
ncbi_id_mapper = kegg_org.kegg_to_ncbi_mapper()
genesets = []
for id in kegg_org.pathways():
pway = kegg.KEGGPathway(id)
hier = ('KEGG', 'pathways')
if pway.pathway_attributes():
kegg_names = kegg_org.get_genes_by_pathway(id)
mapped_genes = []
for gene in kegg_names:
try:
mapped_genes.append(ncbi_id_mapper[gene.upper()])
except KeyError:
# some kegg names can not be matched to ncbi ids
# they are included in geneset anyway
# remove prefix, that specifies kegg organism
# mapped_genes.append(gene.split(':')[-1])
pass
gs = GeneSet(gs_id=id,
name=pway.title,
genes=mapped_genes,
hierarchy=hier,
organism=org,
link=pway.link)
genesets.append(gs)
return GeneSets(genesets)
def kegg_gene_sets(tax_id: str) -> None:
""" Returns gene sets from KEGG pathways.
"""
caching.clear_cache()
kegg_org = kegg.KEGGOrganism(taxonomy.name(tax_id))
ncbi_id_mapper = kegg_org.kegg_to_ncbi_mapper()
genesets = []
for id in kegg_org.pathways():
pway = kegg.KEGGPathway(id)
hier = ('KEGG', 'Pathways')
if pway.pathway_attributes():
kegg_names = kegg_org.get_genes_by_pathway(id)
mapped_genes = set()
for gene in kegg_names:
try:
mapped_genes.add(ncbi_id_mapper[gene.upper()])
except KeyError:
# some kegg names can not be matched to ncbi ids
# they are included in geneset anyway
# remove prefix, that specifies kegg organism
# mapped_genes.append(gene.split(':')[-1])
pass
gs = GeneSet(gs_id=id,
name=pway.title,
genes=mapped_genes,
hierarchy=hier,
organism=tax_id,
link=pway.link)
genesets.append(gs)
for gs_group in GeneSets(genesets).split_by_hierarchy():
hierarchy = gs_group.common_hierarchy()
gs_group.to_gmt_file_format(
f'{data_path}/gene_sets/{filename(hierarchy, tax_id)}')
def get_kgml_and_image(pathway_id):
"""Return an initialized KEGGPathway with pre-cached data"""
p = kegg.KEGGPathway(pathway_id)
p._get_kgml() # makes sure the kgml file is downloaded
p._get_image_filename() # makes sure the image is downloaded
return (pathway_id, p)