Science

Every structural biology course eventually hits the same wall: students can look at a folded protein in PyMOL, they can dock a ligand with AutoDock Vina, but they never actually see the physics doing the work. PyMOL doesn’t visualize forces. Vina’s scoring function is a black box unless you go digging through source code most students will never open. The intuition that molecular dynamics is just Newton’s second law, applied a few million times in a row, gets lost somewhere between the lecture slides and the exercise sheet.

Dataframe operations form the foundation of modern data science workflows, yet researchers often face vendor lock-in when choosing between different dataframe frameworks. My work addresses this by developing a unified operations library that abstracts away framework-specific implementations. The challenge lies in handling the fundamental differences between these systems—some pandas-like dataframes use row-by-row operations, others employ lazy evaluation with native expressions, some require distributed execution patterns, while others focus on parallelization.

Most knowledge about proteins concerns structured proteins with specific functions. However, most proteins don’t have a particular or unique structure, and this doesn’t imply a lack of function. This makes IDPs incredibly interesting. The formation of protein-protein interactions underlies how proteins evolve to perform different functions.