In Mgauth/ccdf: Distribution-Free Single-Cell Differential Expression Analysis

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "README-"
)

`ccdf`

Overview

ccdf is a package for performing single-cell RNA-seq differential expression analysis and more generally complex hypothesis testing.

The main function of the package is ccdf_testing(). It allows to use either an asymptotic test for large sample size or a permutation test for small sample size with the argument method.

The methods implemented in this package are detailed in the following article:

Gauthier M, Agniel D, Thiébaut R & Hejblum BP (2020). Distribution-free complex hypothesis testing for single-cell RNA-seq differential expression analysis, BioRxiv doi:10.1101/2021.05.21.445165

Installation

ccdf is available from CRAN.

install.packages("ccdf")

To install ccdf, you can download the development version on GitHub.

#install.packages("devtools")
devtools::install_github("Mgauth/ccdf")

Example

Here is a basic example which shows how to use ccdf with simple generated data.

## Data Generation
X <- data.frame("X1" = as.factor(rbinom(n=100, size = 1, prob = 0.5)))
Y <- data.frame("gene1" = t(replicate(10, ((X$X1==1)*rnorm(n = 50,0,1)) + ((X$X1==0)*rnorm(n = 50,0.5,1)))))

# Hypothesis testing
res_asymp <- ccdf_testing(exprmat=Y, variable2test=X, test="asymptotic") # asymptotic test
res_perm <- ccdf_testing(exprmat=Y, variable2test=X, test="permutations",
                         adaptive=TRUE) # adaptive permutation test