Prior-Guided Multi-Omic Transformers for Single-Cell Gene Regulatory Network Inference

2026-05-30

AI Summary

This work addresses key limitations in single-cell multi-omics gene regulatory network (GRN) inference—namely, the sparsity of scATAC-seq data, rigid peak-to-gene linkage assumptions, and weak supervision. To overcome these challenges, the authors propose EpiAwareNet, which first employs a gene–peak cross-attention module to enable data-driven, gene-specific aggregation of chromatin accessibility signals. It then leverages somatic GRN priors as noisy positive samples to provide weak supervision under label scarcity, thereby refining regulatory scores. By discarding hard-coded peak–gene mappings and instead adopting adaptive cross-modal representation learning integrated with lightweight prior knowledge, EpiAwareNet substantially outperforms existing methods across multiple benchmarks, significantly improving both the recovery of known regulatory interactions and the biological plausibility of inferred networks.

Abstract

Gene regulatory networks (GRNs) capture transcription factor-target interactions and are central to understanding cell-state regulation and disease. Reconstructing GRNs from paired single-cell transcriptomic and chromatin accessibility data is promising but challenging: scATAC is extremely sparse, and most methods rely on fixed peak-to-gene links and weak supervision. We present EpiAwareNet, a prior-guided multi-omic Transformer framework that reconstructs GRNs from paired single-cell data using only lightweight biological priors. In Stage 1, EpiAwareNet learns joint gene-peak representations with a gene-peak cross-attention module, enabling data-driven, gene-specific aggregation of accessibility signals rather than hard-coded peak-to-gene assignments. In Stage 2, EpiAwareNet incorporates a bulk-derived GRN prior as noisy positive edges to provide weak supervision under label scarcity, refining regulatory scores while remaining robust to prior noise. In our experiments, EpiAwareNet improves GRN reconstruction over representative single- and multi-omic baselines and yields GRNs with greater biological plausibility, such as improved recovery of known regulatory interactions, suggesting that lightweight biological priors from bulk data can effectively guide single-cell GRN inference when combined with adaptive cross-modal representation learning. Code and data will be available at https://github.com/tianyang-x/EpiAwareNet_pub.

Problem

• gene regulatory network
• single-cell multi-omics
• chromatin accessibility
• GRN inference
• sparse data

Innovation

• multi-omic Transformer
• gene regulatory network inference
• cross-attention
• biological priors
• single-cell genomics