griph: Graph Inference of Population Heterogeneity

Panagiotis Papasaikas, Michael Stadler, Atul Sethi
April 2017

griph (Graph Inference of Population Heterogeneity) is an R package for the analysis of single cell RNA-sequencing data. It can be used to automatically identify different cell types or states, even in the presence of confounding sources of variance such as cell cycle stages or batch effects.

System Requirements

In order to use griph, you need:

• R (www.r-project.org, tested with version >= 3.4)
• R packages: igraph, QUIC, coop, corpcor, foreach, doParallel, parallel, bigmemory, RColorBrewer, Rcpp (>= 0.12.9), RcppProgress (>= 0.2.1), RcppArmadillo (>= 0.7.200.2.0)
• A C++ compiler that supports C++11 (e.g. gcc 4.8 or clang 3.3)

To benefit from OpenMP-based parallelization, your compiler must support OpenMP (optional, e.g. gcc 4.8 or clang 4.0)

In order to run griph, especially when using large single cell datasets (more than 5,000 cells), we recommend available memory (RAM) of at least 8 GB.

For a dataset of about 10,000 cells and 20,000 genes, griph takes about three minutes on a standard desktop computer with 4 CPU cores and a total memory of 8 GB.

Installation

griph is free (GPL3) and currently available through https://github.com. It can be installed from R using:

library(devtools)
install_git("git://github.com/ppapasaikas/griph.git", subdir = "griph")
#Or on a windows machine substitute the previous line with:
install_git("https://github.com/ppapasaikas/griph.git", subdir = "griph")

Documentation and Examples

... including step-by-step analysis examples of multiple datasets provided by griph are available from the vignette: (https://ppapasaikas.github.io/griph/)
and from within R:

library(griph)
vignette(griph)