Data Science Seminar
Date: Friday, 09 February 2024
Time: 2:00 PM - 3:00 PM
Venue: NKN Hall
PseudoGA: a pseudo time reconstruction method using single-cell RNA-seq data
Indranil Mukhopadhyay
ISI Kolkata.
09-02-24

---

Abstract

The transition from bulk RNA-seq to single-cell RNA-seq (scRNA-seq) opens tremendous opportunities for deciphering disease-gene architecture. Previously genes were identified as up- or down-regulated by comparing expression data based on bulk RNA-seq that captures only the average effect. This incurs a substantial loss in heterogeneity that is actually present in different cells even from a single sample. Unique technology that generates data from thousands of cells simultaneously is able to capture this heterogeneity. However, this does not depend only on cells but also on the state of the cells, among other factors. Thus it becomes essential to understand the states of the cells. In other words, we should be able to place the cells on a time scale, known as ‘pseudotime’. Dimension reduction-based methods might fail to detect this pseudotime in many situations. We have developed ‘PseudoGA’, a genetic algorithm-based approach using the entire information available in the data. This method seems promising based on our analysis of simulated and real datasets.