Alec Glisman
I am a Senior Machine Learning Scientist at Merck & Co., where I apply predictive deep learning and generative AI to accelerate drug discovery. I bridge physics-based simulation and cheminformatics with machine learning to design, predict, and optimize small molecule therapeutics.
My background spans fluid mechanics, polymer physics, and generative chemistry. I completed my PhD in Chemical Engineering at Caltech in 2024 under Prof. Zhen-Gang Wang. My thesis used molecular simulations of charged polymers and multivalent ions to explain mechanisms of mineralization, adsorption, and polymer-surface interactions. Earlier, I worked with Prof. John F. Brady on self-propulsion in active matter and with Prof. Kranthi K. Mandadapu at UC Berkeley on the mechanics of lipid membranes.
I welcome collaborations in AI for science, molecular modeling, and generative chemistry.
Research Background
My work spans generative AI for drug discovery and the physics of soft matter.
Generative AI, Molecular Design & Property Prediction
AI/MLDrug Discovery
Drug design requires simultaneous optimization of potency, selectivity, pharmacokinetic and safety properties, and synthesizability. I build generative models that jointly optimize these competing objectives while enforcing synthesizability as a hard constraint, so every proposed molecule can actually be made. I also develop ensemble property-prediction models combining graph neural networks with molecular fingerprints, placing 18th out of 400+ submissions in the OpenADMET-ExpansionRx Blind Challenge.
Polyelectrolyte Simulations & Ion Binding
PhysicsAI/ML
Charged polymers (polyelectrolytes) in water bind multivalent ions like calcium, a property exploited in water softening and scale inhibition. Yet the molecular mechanisms of this binding were not well understood. Using all-atom molecular dynamics and enhanced sampling, I showed that indirect ion-ion correlations, not direct bridging between polymer chains, drive chain association. I combined these simulations with unsupervised deep learning to map concentration-dependent phase diagrams and structure-property relationships in collaboration with Dow Chemical.
Education
| PhD, Chemical Engineering | California Institute of Technology, 2024 |
| MS, Chemical Engineering | California Institute of Technology, 2022 |
| BS, Chemical Engineering | University of California, Berkeley, 2019 |