TriTan: An efficient triple non-negative matrix factorisation method for integrative analysis of single-cell multiomics data

Abstract

MotivationSingle-cell multi-omics have opened up tremendous opportunities for understanding gene regulatory networks underlying cell states by simultaneously profiling transcriptomes, epigenomes and proteomes of the same cell. However, existing computational methods for integrative analysis of these high-dimensional multi-modal data are either computationally expensive or limited in interpretation ans scope. These limitations pose challenges in the implementation of these methods in large-scale studies and hinder a more in-depth understanding of the underlying regulatory mechanisms.ResultsHere, we propose TriTan (Triple inTegrative fast non-negative matrix factorisation), an efficient joint factorisation method for single-cell multiomics data. TriTan implements a highly efficient triple non-negative matrix factorisation algorithm which greatly enhances its computational speed, and facilitates interpretation by clustering both the cells and features simultaneously as well as identifying signature feature sets for each cell cluster. Additionally, three matrix factorisation produced by TriTan helps in finding associations of features across modalities, facilitating the prediction of cell type specific regulatory networks. We applied TriTan to single-cell multi-modal data obtained from different technologies and benchmarked it against the state-of-the-art methods where it shows highly competitive performance. Furthermore, we showed a range of downstream analyses that can be conducted utilising the outputs from TriTan.Availabilityhttps://github.com/maxxxxxxxin/TriTanonline.