Xingxin Pan

eSIG-Net predicts how single mutations affect protein interactions using a sequence-based interaction language model, outperforming existing methods.

The paper identifies aberrant R-loops across physiological conditions, develops a deep learning framework to distinguish them from normal R-loops, and links them to cell-specific gene regulation and cancer.

The paper develops a deep learning model integrating multi-omics data to stratify breast cancer patients by prognosis and identify potential therapeutic targets.

D-GPM uses deep learning to predict promoter methylation levels from landmark gene methylation data, outperforming traditional methods in accuracy and correlation.

The paper analyzes genome sequencing data from 60 mammals to infer historical effective population sizes during the Pleistocene, showing how paleoclimate events influenced population fluctuations.