Currently · April 2025 — Present

Building interpretable AI to read the biology hidden inside high-dimensional clinical data.

Under review · Nature-family journals
Transformation of Biological Networks into Images via Semantic Cartography for Visual Interpretation and Scalable Deep Analysis
Nature Computational Science
Under revision (first-round peer review) arxiv.org/abs/2512.07040
Language-Encoded Structural Topology Enables Generalizable Foundation Models for Graph-Structured Data
Nature
Under editorial review arxiv.org/abs/2604.06391
Vision-based Deep Learning Analysis of Unordered Biomedical Tabular Datasets via Optimal Spatial Cartography
Nature Biomedical Engineering
Under external peer review arxiv.org/abs/2603.22675
Research areas
01

Liquid Biopsy & cfRNA

Interpretable models for cancer detection, subtype prediction, and staging across RARE-Seq and independent cfRNA cohorts. Reduced gene panels and biologically grounded attribution maps for clinical adoption.

02

Multimodal Cancer AI

Architectures that integrate molecular omics with medical imaging to surface genetic programs linked to imaging phenotypes.

03

Graph Foundation Models

Graph-to-image and language-encoded topology frameworks that train vision foundation models on biological networks — outperforming GNNs across benchmarks.

04

Explainable AI

XAI techniques for model selection and decision interpretability, spanning plant phenotyping, cancer, and software analytics.

In preparation
  • ·Streamlining Multi-Modal Model for Multi-Omics Integration — Optimal Transport across omics channels.
  • ·Knowledge Graph Modulated Deep Learning for Limited-Sample Clinical Data Analysis — GraphNode classifier with global pathway context.
  • ·A Vision Foundation Model for Biological Networks via Image Transformation of Graph Topology.
  • ·Traditional vs Deep Learning: Benchmarking Aligners and SNP Callers for Low-Coverage Sequencing in Pea.
  • ·Leveraging Feature Explanation for Model Selection: An XAI Framework.
  • ·Genomap Integration Enhances Multi-Omics Visualization and Deep Analysis of Single-Cell Expression Data.