varCluster: Inference of mutational trees by of single cell mutational...

In benstory/mitoClone2: Clonal Population Identification in Single-Cell RNA-Seq Data using Mitochondrial and Somatic Mutations

Inference of mutational trees by of single cell mutational status

Description

From data on the observed mutational status of single cells at a number of genomic sites, computes a likely phylogenetic tree using PhISCS (https://github.com/sfu-compbio/PhISCS) and associates single cells with leaves of the tree. The function clusterMetaclones should be called on the output in order to group mutations into clones using a likelihood-based approach.

Usage

varCluster(
  mutcalls,
  fn = 0.1,
  fp = 0.02,
  cores = 1,
  time = 10000,
  tempfolder = tempdir(),
  python_env = "",
  force_recalc = FALSE,
  method = "SCITE"
)

Arguments

mutcalls	object of class mutationCalls.
fn	false negative rate, i.e. the probability of only observing the reference allele if there is a mutation. #add gene-wise
fp	false positive, i.e. the probability of observing the mutant allele if there is no mutation.
cores	number of cores to use for PhISCS (defaults to 1)
time	maximum time to be used for PhISCS optimization, in seconds (defaults to 10000)
tempfolder	temporary folder to use for PhISCS output
python_env	Any shell commands to execute in order to make the gurobi python package available. The easiest solution is running R from an environment where the gurobi python package is avaiable. In some settings (e.g. RStudio Server), this parameter can be used instead. muta_clone executes PhISCS using a system call to python. The value of this parameter is prepended to the call. If you have a conda environment myenv that contains gurobipy, source activate myenv can work. Occassionally RStudio Server modifies your PATH so that that the conda and source commands are not available. In that case you can for example use export PATH=/path/to/conda/:$PATH; source activate myenv. easybuild users can module load anaconda/v3; source activate myenv
force_recalc	Rerun PhISCS even if the tempfolder contains valid PhISCS output
method	A string variable of either PhISCS or SCITE depending on the tree-inferring software the user wants to use. Default: PhISCS

Value

an object of class mutationCalls, with an inferred tree structure and cell to clone assignment added.

Examples

load(system.file("extdata/LudwigFig7.Rda",package =
"mitoClone2"))
LudwigFig7 <- varCluster(LudwigFig7,
python_env = "",method='SCITE')

benstory/mitoClone2 documentation built on Oct. 30, 2024, 3:20 p.m.