Software & Methods

Neural ordinary differential equations for learning continuous gene-expression dynamics from single-cell data.

DeepVelo: Single-cell transcriptomic deep velocity field learning with neural ordinary differential equations — Science Advances, 2022

• Paper: DeepVelo
• Code: DeepVelo

Gradient-boosting models that connect cis-regulatory elements to target genes using single-cell multiomics.

DIRECT-NET: An efficient method to discover cis-regulatory elements and construct regulatory networks from single-cell multiomics data — Science Advances, 2022

• Paper: DIRECT-NET
• Code: DIRECT-NET

Spatial gene-expression imputation using graph paths that integrate spatial proximity and expression similarity.

Impeller: a path-based heterogeneous graph learning method for spatial transcriptomic data imputation — Bioinformatics, 2024

• Paper: Impeller
• Code: Impeller

Joint analysis of spatial transcriptomic slices through optimal-transport alignment and graph representation learning.

MUSE: A Multi-slice Joint Analysis Method for Spatial Transcriptomics Experiments — CIKM, 2025

• Paper: MUSE
• Code: MUSE

Differentiable combinatorial optimization for prioritizing candidate causal variants in the non-coding genome.

MUGO: Differentiable Combinatorial Optimization for Causal Variant Discovery in the Non-coding Genome — KDD, 2026

Reparameterizable subset explanations for identifying combinations of regulatory genes associated with cell-fate decisions.

Identifying Combinatorial Regulatory Genes for Cell Fate Decision via Reparameterizable Subset Explanations — KDD, 2025

• Paper: MetaVelo
• Code: MetaVelo

Iterative multi-view graph learning to model intercellular gene regulation from spatial transcriptomic data.

iMIRACLE: An Iterative Multi-View Graph Neural Network to Model Intercellular Gene Regulation from Spatial Transcriptomic Data — CIKM, 2024

• Paper: iMIRACLE
• Code: iMIRACLE

Hierarchical graph representations for measuring network rewiring and prioritizing disease-associated genes.

iHerd: an integrative hierarchical graph representation learning framework to quantify network changes and prioritize risk genes in disease — PLOS Computational Biology, 2023

• Paper: iHerd
• Code: iHerd