Multi-scale single-cell clustering (MSC) workflow

main @ b8c516e

Overview

Developmental version of MSC: This github page contains developmental version of R package for Multi-scale clustering (MSC) to perform single-cell transcriptome clustering. The manuscript is currently under review.

Installation:

MEGENA needs to be installed, prior to MSC installation: library(devtools); install_github("songw01/MEGENA");

For installation for developmental github version:
library(devtools);
install_github("songlabcodes/MSC");

Vignettes
PBMC 8k example: This contains the overall MSC workflow from data pre-processing, clustering, marker identification to visualization of the results.

pipeComp vignettes: Comparative analysis of MSC performance with other benchmark clustering methods using pipeComp data sets with gold standard ground-truth clusters. Reproduces Figure 4 of manuscript.

Ding et al. 2020 vignettes: Comparative analysis of MSC performance with other benchmark clustering methods using pipeComp data sets with gold standard ground-truth clusters. Reproduces Figure 5 of manuscript.

Lee et al. 2020 vignette: Codes to reproduce MSC analysis of influenza/COVID-19 infected PMBC scRNA-seq from Lee et al. 2020.

Wu et al. 2021 vignette: Codes to reproduce MSC analysis of breast cancer single-cell atlas from Wu et al. 2021.

Disclaimer: Data files for examples (pbmc_8k_msc_results.rda, simMix1.rda) are distributed under the CC0 1.0 Universal (CC0 1.0) Public Domain Dedication (https://creativecommons.org/publicdomain/zero/1.0/). Otherwise, the codes are distrubuted under GPL-3.0 license as creative works.