Integrating multiple single-cell multi-omics samples with Smmit

Changxin Wan,Zhicheng Ji

DOI: https://doi.org/10.1101/2023.04.06.535857

2024-12-10

Abstract:Multi-sample single-cell multi-omics datasets, which simultaneously measure multiple data modalities in the same cells across multiple samples, facilitate the study of gene expression, gene regulatory activities, and protein abundances on a population scale. We developed Smmit, a computational method for integrating data both across samples and modalities. Compared to existing methods, Smmit more effectively removes batch effects while preserving relevant biological information, resulting in superior integration outcomes. Additionally, Smmit is more computationally efficient and builds upon existing computational pipelines, requiring minimal effort for implementation. Smmit is an R software package that is freely available on Github: https://github.com/zji90/Smmit.

Bioinformatics