10X based Single cell sequencing for comparing two different disease states

Question

[halo22]

We are thinking about designing an experiment comparing disease profiles of individuals belonging to two different disease states, the idea would be compare gene expression profiles from using T-cells. Since the goal is comparison of gene expression profiles, what platform would you recommended for a such study? Has anyone used 10X genomics single cell platform to design a similar study? As single cell data is zero-inflated what tools should be used to run the DGE analysis?

Answer 1

[igor]

This is a very common type of analysis. There is a recent paper by Soneson & Robinson with an overview of the challenges and the available methods:

One of the most commonly performed tasks for RNA-seq data is differential gene expression (DE) analysis. Although well-established tools exist for such analysis in bulk RNA-seq data, methods for scRNA-seq data are just emerging. Given the special characteristics of scRNA-seq data, including generally low library sizes, high noise levels and a large fraction of so-called 'dropout' events, it is unclear whether DE methods that have been developed for bulk RNA-seq are suitable also for scRNA-seq. A few recent studies suggest that the optimal method may depend on the number of cells and strength of the signal and that methods not initially developed for scRNA-seq analysis can perform well. In this study, we used processed data sets, from conquer and other sources, to evaluate DE methods in scRNA-seq data. Our study expands the number of methods and range of experimental data sets assessed in previous comparisons and includes evaluations based on simulated data. We also investigated the effect of filtering out lowly expressed genes and extended the set of evaluation criteria.