AUTHOR=Ellis Dorothy , Roy Arkaprava , Datta Susmita 

TITLE=Clustering single-cell multimodal omics data with jrSiCKLSNMF

JOURNAL=Frontiers in Genetics

VOLUME=Volume 14 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2023.1179439

DOI=10.3389/fgene.2023.1179439

ISSN=1664-8021

ABSTRACT=The development of multi-assay single cell omics methods has enabled the collection of data across different omics views from the same set of single cells. Each omics view provides unique data about cell type and function, so the ability to integrate data from different views can provide deeper insights into cellular functions. Often, single cell omics data can prove challenging to model because of high dimensionality, sparsity, and technical noise. We propose a novel multi-assay data analysis method, joint graph-regularized Single-Cell Kullback-Leibler Sparse Non-negative Matrix Factorization (jrSiCKLSNMF, pronounced ”junior sickles NMF”) that extracts latent factors shared across omics views within the same set of single cells. Furthermore, we develop multiplicative updates based efficient computation and show overwhelmingly better clustering performance than several existing methods on data simulated from third-party software. On a real multi-assay omics dataset, we also find our method to produce scientifically accurate clustering results.