Inferring differential dynamics from multi-lineage, multi-omic, and multi-sample single-cell data with MultiVeloVAE.

ABSTRACT: Understanding how cells differentiate to their final specialized fates is a fundamental problem in biomedical science. Single-cell multi-omic profiling provides an opportunity to identify dynamic molecular changes, but new computational approaches are needed to realize this potential. In particular, previous methods for RNA velocity inference lack support for multi-lineage, multi-sample, and multi-omic single-cell data and cannot be used to identify differential dynamics. To overcome these challenges, we introduce MultiVeloVAE, a probabilistic framework for multi-sample RNA velocity inference that integrates single-cell RNA and multi-omic data. MultiVeloVAE models gene expression and chromatin accessibility on a shared time scale, performs multi-sample inference from datasets with partiall

SUBMITTER: Li C 

PROVIDER: S-EPMC12748740 | biostudies-literature | 2025 Nov

REPOSITORIES: biostudies-literature