find_subclones: Identify subclones in a clonal branch of a hierarchical tree.

In seqpipe/SCclust: Clustering of Single Cell Sequencing Copy Number Profiles to Identify Clones

Description

Iterate the procedure for clone identification for a subset of cells forming a clone.

Usage

find_subclones(hc, pinmat, pins, nmin = 6, nsim = 500, njobs = NULL,
  lmmax = 0.001, hcmethod = "average", baseshare = 3, fdrthresh = -2,
  sharemin = 0.85, bymax = T, climbfromsize = 2, climbtoshare = 3,
  clonetype = "soft")

Arguments

hc	The hclust object with clones identified.
pinmat	The feature incidence matrix: columns are cells, rows are features, 1 if a feature is present, 0 if not.
nmin	An integer. Default: 6. The minimal allowed size of a clone to be examined for subclones.
nsim	The number of permutation simulations for subclone identification. Default: 500.
lmmax	Numeric value. Default: 0.001. The threshold parameter for a linear fit, passed to fisherfdr function.
hcmethod	Default: average
baseshare	An integer. Default: 3. A balance parameter for controlling minimal number of shared features in a subclone node.
fdrthresh	FDR criterion for subclone nodes. Default: -2.
sharemin	A feature is considered shared if present in sharemin fraction of leaves in a node.Default: 0.85.
bymax	Logical. If TRUE (Default), use maximal pairwise FDR for the node to find subclones, otherwise use mean over all pairs.
climbfromsize	An integer specifying the minimal size of a hard subclone allowed to be expanded. Default: 2.
climbtoshare	An integer the minimal number of widely shared features in a soft subclone Default: 3.
A	two-column matrix, one row per feature, providing the bin number and thepy (sign) of the feature.
clonetype.	A character string specifying whether hard or soft subclones are to be determined. Default: 'soft'.

Value

A list of hclust objects for clones.

seqpipe/SCclust documentation built on May 4, 2021, 11:56 a.m.