BAF_EM_VAF: BAF_EM_VAF

In pllittle/UNMASC: Tumor-Only Variant Calling with Unmatched Normal Control Samples

BAF_EM_VAF

Description

Clusters integer read counts to model mixtures of noise distributions, binomial vs beta-binomial distributions, over-dispersion.

Usage

BAF_EM_VAF(MAT, binom = TRUE, show_dots = TRUE, clust_tVAF = NULL)

Arguments

MAT	A numeric matrix of alternate (AD) and reference (RD) read counts in addition to pre-calculated total read depth DP = (AD + RD), log-transformed binomial coefficient (LBC), log-transformed total read depth (log_DP).
binom	Boolean set to true by default to model the mixture distribution assuming binomial distribution. Otherwise set to false to explore beta-binomial and binomial models.
show_dots	Boolean set to true by default to visualize computational runtime.
clust_tVAF	Boolean but set to null to indicate clustering normal read counts represented by modeling a mixture of noise and/or VAF cluster centered around 0.5. Otherwise when set to true, the function models the mixture by noise, cluster at 0.5, clusters p and 1-p for copy-altered genomic segments.

Value

A list of clustered parameter estimates and posterior probabilities for inferring classification.

pllittle/UNMASC documentation built on June 1, 2025, 1 p.m.