A0583
Title: Modelling single-cell RNA-seq data
Authors: Pronoy Kanti Mondal - Johns Hopkins University (United States)
Indranil Mukhopadhyay - University of Nebraska - Lincoln (United States) [presenting]
Abstract: Analysis of single-cell RNA-seq data is challenging due to severe sparsity and the influence of many interacting biological factors. Any downstream analysis requires appropriate modeling of raw data, taking care of the inherent complexities present in the data. Bimodal patterns of the expression data add more complexity. The aim is to propose a statistical model that takes care of all these factors simultaneously and fits a probability distribution on the expression levels of each gene, leveraging information from the entire data set. Based on this modeling, a two-sample testing method is also developed for testing differential expression between two groups. Extensive data analysis is performed based on simulated and real data sets to validate the methods.
