Topic modelling in millions of single-cell genomics profiles without compromising high-dimensionality
Conference
64th ISI World Statistics Congress
Format: IPS Abstract
Keywords: genomics, matrix
Session: IPS 230 - Recent Developments in Statistical Genetics and Genomics
Thursday 20 July 2 p.m. - 3:40 p.m. (Canada/Eastern)
Abstract
We propose a simple idea to speed up probabilistic topic modelling in millions of single-cell data. We show that our method termed ASAP performs as well as other methods but requires much fewer computing resources and run time. We applied our method to massive single-cell expression data sets and uncovered rare cell types and gene regulatory programs.