reDA: differential abundance testing on scATAC-seq data using random walk with restart.

Abstract

Identifying cell states associated with disease progression or experimental perturbations from single-cell Assay for Transposase Accessible Chromatin using sequencing (scATAC-seq) data is critical for unraveling disease pathogenesis. However, the high dimensionality, extreme sparsity, and nearly binary nature of scATAC-seq data pose significant challenges. Here, we present reDA, a cluster-free computational framework that performs differential abundance testing based on the random walk with restart. Through comprehensive experiments on simulated and real datasets, reDA outperforms six baseline methods, demonstrating superior accuracy, computational efficiency, and the ability to capture disease-specific molecular signatures. The reDA along with detailed documentation is freely available at https://github.com/Jinsl-lab/reDA. It can be seamlessly integrated into existing scATAC-seq analysis workflows.