scGPL framework unifies graph and pairwise learning for gene networks
November 11, 2026
The scGPL framework addresses overfitting in graph-based methods and topological blindness in pairwise-based methods for gene regulatory network inference. It jointly models global regulatory topology and individual TF-target gene pairs using scRNA-seq data.
HOW THIS AFFECTS YOU
●
researcherYou can use this end-to-end framework to get more accurate gene interaction models from single-cell data.
●
healthThis may improve how researchers understand the underlying regulatory mechanisms in complex biological systems.