mut_example_multi: Posterior sampling chains with cancer mutation data

In nicolaroberts/hdp: Hierarchical Dirichlet Process for categorical count data

Description

Four independent HDP sampling chains with data from SomaticCancerAlterations package, saved to mut_count. Categories are the 96 base substitution types defined by local trinucleotide content, and the samples include 100 lung adenocarcinomas, 100 ovarian serous carcinomas, and 100 skin cutaneous melanomas. Data is derived from TCGA exome-sequencing studies. Each sample was assigned to a unique child DP node, with one parent DP node per cancer type, and one grandparent DP node at the top level. Each chain initialised with 10 clusters, then run 5000 burn-in iterations before collecting 50 posterior samples with 200 iterations between each.

Usage

mut_example_multi

Format

A hdpSampleMulti object with 200 posterior samples, 50 from each chain

nicolaroberts/hdp documentation built on May 23, 2019, 5:09 p.m.