BATMAN: Fast and Accurate Integration of Single-Cell RNA-Seq Datasets via Minimum-Weight Matching

• DOI: 10.1016/j.isci.2020.101185
• Publication Date: 2020-05-20T17:19:35Z
• Authors: Igor Mandric; Brian L. Hill; Malika K. Freund; Michael Thompson; Eran Halperin
• Keywords: 0301 basic medicine; 570; 0303 health sciences; Bioinformatics; Science; Human Genome; Q; Bioinformatics and Computational Biology; Biological Sciences; Article; 004; 03 medical and health sciences; Genetics; Transcriptomics; Algorithms; Biotechnology

Single-cell RNA-sequencing (scRNA-seq) is a set of technologies used to profile gene expression at the level individual cells. Although throughput scRNA-seq experiments steadily growing in terms number cells, large datasets are not yet commonly generated owing prohibitively high costs. Integrating multiple into one can improve power experiments, and efficient integration very important for downstream analyses such as identifying cell-type-specific eQTLs. State-of-the-art methods based on mutual nearest neighbor paradigm fail both correct batch effects maintain local structure datasets. In this paper, we propose novel dataset method called BATMAN (BATch via minimum-weight MAtchiNg). Across simulations real datasets, show that our significantly outperforms state-of-the-art tools with respect existing metrics by up 80% while retaining cell-to-cell relationships.