scCNMF: an integrated analysis model for paired single-cell RNA sequencing and assay for transposase-accessible chromatin sequencing data leveraging cell similarity and cis-regulatory potential.

The integrated analysis of paired single-cell RNA sequencing (scRNA-seq) and single-cell Assay for Transposase-Accessible Chromatin using sequencing (scATAC-seq) data is crucial for accurately characterizing cellular states and reconstructing gene regulatory networks. However, most integration methods fail to simultaneously consider the high sparsity of scATAC-seq data and regulatory interactions at the cellular level, limiting the biological interpretability and accuracy of their integration results. In this study, we present scCNMF, a novel model for the integrated analysis of paired scRNA-seq and scATAC-seq data. scCNMF based on the non-negative matrix factorization model, jointly incorporates cell similarity structures and prior regulatory information, leading to improved cell embeddings and enhanced clustering accuracy. We evaluate scCNMF on multiple real-world datasets and demonstrate that it achieves competitive performance compared to state-of-the-art methods. Moreover, further analyses show that scCNMF enables biomarker identification, highlighting its interpretability. Additionally, scCNMF facilitates signal enhancement of scATAC-seq data, resulting in improved data quality for subsequent analyses.