mSigHdp: Mutational Signature Discovery Using Hierarchical Dirichlet Processes

The goal of mSigHdp is mutational signature discovery using hierarchical Dirichlet process (HDP) mixture models. mSigHdp is only supported on Linux systems. Most users will use the function RunHdpxParallel.

This package uses https://github.com/steverozen/hdpx for the hierarchical Dirichlet process implementation.

Please also see our paper: Mo Liu, Yang Wu, Nanhai Jiang, Arnoud Boot, Steven G. Rozen, mSigHdp: hierarchical Dirichlet process mixture modeling for mutational signature discovery, https://doi.org/10.1093/nargab/lqad005.

Installation

Newest stable version

Compatible with current Rcpp. Corrects Error in lower.to.upper.tri.inds(n) : 'n' must be >= 2

install.packages("remotes")
remotes::install_github(repo = "steverozen/mSigHdp", ref = "v2.1.3-branch")

Very Important after intallation make sure that `packageVersion('hdpx') returns "1.0.6". If not, do

remotes::install_github(repo = "steverozen/hdpx", ref = "v1.0.6-branch")

Singularity/Apptainer

This is v2.1.2 and still has ``Error in lower.to.upper.tri.inds(n) : 'n' must be >= 2`
on inputs with only a single signature.

singularity pull library://rozen-lab/msighdp/msighdp:2.1.2

For more details, see https://github.com/steverozen/mSigHdp/blob/master/data-raw/container_scripts/mSigHdp_container_installation.pdf

A toy-example R script for using this container is available at https://github.com/steverozen/mSigHdp/blob/v2.1.2-branch/data-raw/container_scripts/test_mSigHdp.R.

Previous stable version

Incompatible with current Rcpp

install.packages("remotes")
remotes::install_github(repo = "steverozen/mSigHdp", ref = "v2.1.2-branch")

Reference manual

https://github.com/steverozen/mSigHdp/blob/v2.1.2-branch/mSigHdp_2.1.2.pdf