Integrating pathways and related knowledge in a multilayer framework

Automatic context-specific network inference.

A matrix factorization model for multi-omic data and biological pathways

Analysis pipeline for single-cell RNA sequencing (scRNA-seq) data focused on T cells in COVID-19. Includes scripts and workflows for data processing, visualization, and identifying immune cell signatures and responses. Designed to accelerate reproducible research in immunology and single-cell genomics.

SpatialTranscriptFormer: A biologically-informed transformer architecture for spatial transcriptomics. Features Quad-Flow Interaction (P↔P, P↔H, H→P, H↔H) and a pathway bottleneck using MSigDB Hallmarks. Integrated with pathology foundation models (CTransPath/Phikon) and optimized via composite MSE+PCC loss to ensure spatial coherence.