Name Mode Size
R 040000
data 040000
man 040000
tests 040000
vignettes 040000
.Rbuildignore 100644 0 kb
.gitignore 100644 0 kb
CITATION.cff 100644 1 kb
DESCRIPTION 100644 1 kb
LICENSE 100644 1 kb
NAMESPACE 100644 0 kb 100644 1 kb
README.Rmd 100644 3 kb 100644 2 kb
<!-- is generated from README.Rmd. Please edit that file --> # DCATS <!-- badges: start --> <!-- badges: end --> This R package contains methods to detect the differential composition abundances between multiple conditions in singel-cell experiments. The **latest** version of the `DCATS` package is 0.99.7. It is under the MIT license. ## Installation ### From R The **latest** `DCATS` package can be conveniently installed using the [`devtools`]( package thus: ``` r ## install dependencies install.packages(c("MCMCpack", "matrixStats", "robustbase", "aod", "e1071")) ## dependencies for vignette install.packages(c("SeuratObject", "Seurat", "robustbase", "aod", "e1071")) devtools::install_github('satijalab/seurat-data') ``` ``` r # install.packages("devtools") devtools::install_github("holab-hku/DCATS", build_vignettes = TRUE) ``` You can also install `DCATS` without building the vignette: devtools::install_github("holab-hku/DCATS") ### From Biocounductor (required R &gt;= 4.3.0) ``` r if (!requireNamespace("BiocManager")) install.packages("BiocManager") BiocManager::install("DCTAS") ``` #### For development Download this repository to your local machine and open it in Rstudio as a project, and build it by install and restart. ## Getting started The best place to start are the vignettes. Inside an R session, load `DCATS` and then browse the vignette about the usage guidance of `DCATS`: ``` r library(DCATS) browseVignettes("DCATS") ``` The tutorial demonstrating how to use DCATS after using [`Seurat`]( pipeline to process data can be found in [`Integrate DCATS with Seurat pipeline`]( ### Example This is a basic example which shows you how to estimate a similarity matrix from KNN graph and do the differential abundance test using this similarity matrix. ``` r library(DCATS) data("simulation") knn_mat = knn_simMat(simulation$knnGraphs, simulation$labels) sim_count = rbind(simulation$numb_cond1, simulation$numb_cond2) sim_design = data.frame(condition = c("c1", "c1", "c2")) knn_mat[colnames(sim_count),] res = dcats_GLM(as.matrix(sim_count), sim_design, similarity_mat = knn_mat) print(res$LRT_pvals) ```