clone_cluster: clone_cluster

In mskilab/skitools: Various mskilab R utilties

clone_cluster

Description

determines "clone membership" using CCF threshold for n variants, given alt read count, total read count, and purity

first fits k component mixture model (k pre-specified) to CCF histogram .. mixture model can also be given as input returns cluster centers and membership probabilities for each mutation

$mu k vector of means $sigma k vector of sigma $p.cluster n x k matrix of cluster probabilities $lambda k vector of cluster membership frequencies return n x k matrix of probabilities that CCF>ccf.thresh for each variant

altc, totc, and cn are of length n, purity is length 1

Usage

clone_cluster(
  altc,
  totc,
  cn,
  purity,
  thresh = 0.95,
  k = 2,
  mix.model = NULL,
  grid.size = 0.01,
  verbose = F,
  nsamp = 10000
)

Author(s)

Marcin Imielinski

mskilab/skitools documentation built on Aug. 31, 2023, 1:13 p.m.