Feng Gao

The paper develops eTRex, a variational Bayesian model, to analyze 4,819 ATAC-seq datasets and create a pan-cancer atlas of transcriptional regulator profiles, revealing common and context-specific regulators.

Neurospectrum extracts interpretable spatiotemporal features from neural activity using a geometric and topological deep learning framework, outperforming traditional methods in simulations and fMRI data.