fetch gene expression and mutation from Pancan Atlas cohort

pancan_example.py

import xenaPython as xena

host = xena.PUBLIC_HUBS["pancanAtlasHub"]
cohort ="TCGA PanCanAtlas"
samples = xena.cohort_samples(host, cohort, None)

genes = ["FOXM1", "TP53"]
expression_dataset = "EB++AdjustPANCAN_IlluminaHiSeq_RNASeqV2.geneExp.xena"
[foxm1_expresssion_by_sample, tp53_expression_by_sample] = xena.dataset_fetch(host, expression_dataset, samples, genes)

mutation_dataset = "mc3.v0.2.8.PUBLIC.xena"
mutations = xena.sparse_data(host, mutation_dataset, samples, genes)
# mutation data is returned in column orientation, i.e. ['rows'] is a dict of
# arrays, one per column, each having length N, where N is the length of the
# result set.
all_rows = range(len(mutations['rows']['sampleID']))
foxm1_rows = [i for i in all_rows if mutations['rows']['genes'][i][0] == 'FOXM1']
tp53_rows = [i for i in all_rows if mutations['rows']['genes'][i][0] == 'TP53']


phenotype_dataset = "Survival_SupplementalTable_S1_20171025_xena_sp"
subtype_field = "cancer type abbreviation"
cancer_subtype = xena.dataset_fetch(host, phenotype_dataset, samples, [subtype_field])[0]
codes = xena.field_codes(host, phenotype_dataset, [subtype_field])[0]['code'].split('\t')

cancer_subtype_by_sample = [None if v == 'NaN' else codes[int(v)] for v in cancer_subtype]