Ahmad Ali

Hi! I’m Ahmad Ali, completing my B.Sc. at Toronto Metropolitan University (August 2026). I am a research assistant with the Biomathematics and Fluids Group under the supervision of Dr. Kathleen Wilkie.

My Vision

I am motivated by the challenge of decoding molecular complexity. Solving the fundamental problems of proteomics, material science, and digital twin biology requires a paradigm shift—moving beyond deterministic approximations toward probabilistic AI architectures capable of modeling stochastic systems with high accuracy.

Research

My work bridges mechanistic modeling and deep learning. In computational biology, I develop predictive models in Julia, utilizing solver optimization and Markov Chain Monte Carlo (MCMC) simulations to quantify uncertainty in viral dynamics.

Simultaneously, I investigate AI for Science. My recent work on Machine Learning Interatomic Potentials (MLIPs) challenges the standard regression paradigm used in foundation models like UMA. I demonstrated that training with histogram-based cross-entropy loss enables the quantification of epistemic uncertainty—a critical feature absent in traditional scalar regression.

I focus my work by constantly interrogating the science behind it—pulling apart papers, tracing methods, and presenting ideas until the assumptions reveal themselves. That process is what hooked me. The gaps in our models, and the architectures still waiting to be invented, are what keep me pushing forward.