Research

My interests lie at the interface of technology with (bio)pharma and healthcare. In the past, I developed and applied machine learning and physics-based methods for drug, protein, and polymer design; below are some examples.

Molecular property prediction

Leveraging computation for the design of organic and biological molecules often means trying to predict (bio)molecular properties. I've worked on predicting protein-ligand binding affinity with molecular dynamics simulations (Chem. Sci. 2021, Chem. Sci. 2020, JACS 2017, Chem. Sci. 2016), and how affinity changes as a consequence of protein mutation, with both simulations and ML (ACS Cent. Sci. 2018, ACS Cent. Sci. 2019). I developed graph neural networks tailored to polymer property prediction (Chem. Sci. 2022), and ML models for the design of polymeric long-acting injectables (Nat. Commun. 2023).

Active learning and design of experiment

Numerous tasks in science and engineering can be framed as multi-objective optimization problems. I've spent some time working on software and algorithms for model-guided optimization (Acc. Chem. Res. 2021, Mach. Learn. Sci. Technol. 2021). In particular, I've focused on developing or tailoring data-driven approaches to suit typical scenarios encountered in experimetal chemistry and materials science, such as categorical (Appl. Phys. Rev. 2021), constrained (Dig. Discov. 2022), and robust (Chem. Sci. 2021) optimization. When combined with automated hardware, these approaches can enable research platform capable of autonomous experimentation ("self-driving" laboratories; Adv. Funct. Mater. 2021).

Biomolecular mechanisms and dynamics

Computer simulations allow one to study biochemical systems in atomistic detail, exploring the multitude of conformations available to molecules in solution, and extracting information about their thermodynamic, kinetic, and dynamical properties. Ultimately, they provide insight into the molecular determinants of biological function and dysfunction. In the past, I've used simulations to study the role of water molecules in mediating protein-ligand binding (Sci. Adv. 2015, Commun. Chem. 2018), to propose a molecular mechanism for transthyretin amyloidogenesis (Nat. Commun. 2019), and to decipher the structural basis for the antibiotic action of antivitamins (Nat. Chem. Biol. 2020).