Gurpreet Bharj

Rapid and Interpretable AMR Diagnostics via Genomics and Cell Painting using Differential Geometry-based Directed-Simplicial Neural Networks on Multimodal Data

This paper develops a multimodal AI model for rapid AMR prediction using genomic and cellular data, identifying key genomic motifs and phenotypic features associated with antimicrobial resistance.

MAP-PRS: Multi-Ancestry Portfolio-Based Polygenic Risk Scores

MAP-PRS improves polygenic risk score fairness and reliability across populations by combining ancestry-specific models with portfolio optimization and computational tools.

Interpretable Machine Learning and Comparative Genomics Reveal Microbial Plastic-Degrading (Microbeyt) Potential

The paper identifies microbial strains with plastic-degrading potential using a computational pipeline combining genomic analysis and interpretable machine learning.