CHAI: consensus Clustering tHrough similArIty matrix integratIon for single cell type identification

Latest News

Version 0.99.0 (2023-03-15)

• Initial release of the package.
• Submitted to Bioconductor.

For the full log of news and updates, please check the NEWS.md file.

Introduction

CHAI (consensus Clustering tHrough similArIty matrix integratIon for single cell type identification) is a consensus clustering framework that offers two methods for consensus clustering: Average Similarity (AvgSim) and Similarity Network Fusion (SNF) (Wang et al. 2014).

Installation Instructions

You may install CHAI using devtools.

devtools::install_github("lodimk2/chai")

Quick Start

CHAI contains two methods for consensus clustering: CHAI-AvgSim and CHAI-SNF. We provide wrapper functions to run either of these.

We provide example data from the Baron Mouse 1 Dataset (Baron et al., 2016). Data should be in the form of a SingleCellExperiment object, with counts and logcounts defined.

Load Dependencies:

library(chai)
library(scSHC)
library(RaceID)
library(SC3)
library(SingleCellExperiment)
library(CHOIR)

Load Data:

data("baron_mouse_1")
# Create SingleCellExperiment object
sce <- SingleCellExperiment(assays = list(counts = as.matrix(baron_mouse_1)))
# Add logcounts 
sce <- scuttle::logNormCounts(sce)

CHAI-AvgSim:

# If "eval" is set to TRUE, CHAI will evaluate the best_k for Spectral Clustering using silhouette score. If "eval" is set to false, CHAI will use best_k as the k value for silhouette score.
# The default is eval being set to TRUE
best_k <- 15
sce <- CHAI_AvgSim <- function(sce,best_k,eval = TRUE)

CHAI-SNF:

# If "eval" is set to TRUE, CHAI will evaluate the best_k for Spectral Clustering using silhouette score. If "eval" is set to false, CHAI will use best_k as the k value for silhouette score.
# The default is eval being set to TRUE
best_k <- 15
sce <- CHAI_SNF <- function(sce,best_k,eval = TRUE)

In Depth Tutorial

For a more detailed tutorial, including adding your own clustering algorithms to CHAI and including other "omics" data, please see the vignette inst/chai.html.

References:

1. Baron M, Veres A, Wolock SL, Faust AL, Gaujoux R, Vetere A, Ryu JH, Wagner BK, Shen-Orr SS, Klein AM, Melton DA, Yanai I. A Single-Cell Transcriptomic Map of the Human and Mouse Pancreas Reveals Inter- and Intra-cell Population Structure. Cell Syst. 2016 Oct 26;3(4):346-360.e4. doi: 10.1016/j.cels.2016.08.011. Epub 2016 Sep 22. PMID: 27667365; PMCID: PMC5228327.
2. Wang, B., Mezlini, A., Demir, F. et al. Similarity network fusion for aggregating data types on a genomic scale. Nat Methods 11, 333–337 (2014). https://doi.org/10.1038/nmeth.2810

Citation:

If you use CHAI, please cite our manuscript:

Lodi, M., Lodi, M., Osei, K., Ranganathan, V., Hwang, P., & Ghosh, P. (2024). CHAI: Consensus Clustering Through Similarity Matrix Integration for Cell-Type Identification. bioRxiv. https://doi.org/10.1101/2024.03.19.585758

@article{CHAI_BIB,
  title = {CHAI: Consensus Clustering through Similarity Matrix Integration for Cell-Type Identification},
  author = {Musaddiq K. Lodi and Muzammil Lodi and Kezie Osei and Vaishnavi Ranganathan and Priscilla Hwang and Preetam Ghosh},
  journal = {Briefings in Bioinformatics},
  volume = {25},
  number = {5},
  year = {2024},
  doi = {10.1093/bib/bbae411},
  publisher = {Oxford University Press},
}

Questions and Support:

For any questions or comments, please reach out to Musaddiq Lodi @ lodimk2@vcu.edu