Single-Cell RNA-Seq Gene Regulatory Network Inference via Explicit Link Modeling and Adaptive Architectures

This release provides the first public, reproducible implementation of AutoGERN, a link-centric graph neural network framework for gene regulatory network (GRN) inference from scRNA-seq. AutoGERN explicitly learns edge (link) embeddings in message passing and integrates a stochastic differentiable architecture search to adapt GNN operators to dataset-specific distribution shifts.