Recent work has enabled the simultaneous analysis of RNA levels and chromatin accessibility for gene regulatory network (GRN) inference, but heterogeneity across omics layers remains a barrier in these approaches. Here, the authors present scTFBridge, a deep-generative model for GRN inference, whereby latent spaces are disentangled into shared and specific components across omics layers.
- Feng-ao Wang
- Chenxin Yi
- Yixue Li
