Prior-Guided Multi-Omic Transformers for Single-Cell Gene Regulatory Network Inference
Researchers have developed EpiAwareNet, a novel framework utilizing multi-omic Transformers to infer gene regulatory networks (GRNs) from single-cell data. This method integrates transcriptomic and chromatin accessibility information, overcoming challenges like data sparsity and reliance on fixed gene-peak links. EpiAwareNet employs a gene-peak cross-attention module for adaptive signal aggregation and incorporates a prior GRN from bulk data as weak supervision, enhancing biological plausibility and improving reconstruction accuracy over existing methods. AI
IMPACT This new method could advance our understanding of cell regulation and disease by improving the accuracy of gene regulatory network reconstruction.