BSDE: barycenter single-cell differential expression for case-control studies

Motivation Single-cell sequencing brings about a revolutionarily high resolution for finding differentially expressed genes (DEGs) by disentangling highly heterogeneous cell tissues. Yet, such analysis is so far mostly focused on comparing between different cell types from the same individual. As single-cell sequencing becomes cheaper and easier to use, an increasing number of datasets from case-control studies are becoming available, which call for new methods for identifying differential expressions between case and control individuals. Results To bridge this gap, we propose barycenter single-cell differential expression (BSDE), a nonparametric method for finding DEGs for case-control studies. Through the use of optimal transportation for aggregating distributions and computing their distances, our method overcomes the restrictive parametric assumptions imposed by standard mixed-effect-modeling approaches. Through simulations, we show that BSDE can accurately detect a variety of differential expressions while maintaining the type-I error at a prescribed level. Further, 1345 and 1568 cell type-specific DEGs are identified by BSDE from datasets on pulmonary fibrosis and multiple sclerosis, among which the top findings are supported by previous results from the literature. Availability and implementation R package BSDE is freely available from doi.org/10.5281/zenodo.6332254. For real data analysis with the R package, see doi.org/10.5281/zenodo.6332566. These can also be accessed thorough GitHub at github.com/mqzhanglab/BSDE and github.com/mqzhanglab/BSDE_pipeline. The two single-cell sequencing datasets can be download with UCSC cell browser from cells.ucsc.edu/?ds=ms and cells.ucsc.edu/?ds=lung-pf-control. Supplementary information are available at Bioinformatics online.

Automated summary

Single-cell sequencing has revolutionized the identification of differentially expressed genes (DEGs) by providing high resolution for heterogeneous cell tissues. However, current analysis mainly focuses on comparing different cell types within the same individual. This study proposes a nonparametric method called barycenter single-cell differential expression (BSDE) to identify DEGs in case-control studies. The method overcomes the limitations of parametric approaches and accurately detects differential expressions. It is demonstrated through simulations and real data analysis that BSDE can effectively identify cell type-specific DEGs. The availability of the R package and datasets further facilitate its application in research.