Microhydrodynamics to Pharmaceutical Informatics:

A Journey of Discovery in Computational Modeling

2013 Ho-Am Commemorative Lecture in Engineering

May – June 2013

Sangtae Kim

Abstract

“Every student of science and engineering knows that a particle in a vacuum subject to an external force undergoes an acceleration given by Newton’s famous law. The particle trajectory is obtained immediately upon two integrations, and thus the fate of the particle can be determined. Every student of fluid dynamics learns that in a fluid medium, viscous resistance of the fluid decreases this acceleration so that the particle soon attains a terminal velocity … drag is shape-dependent.”

[Opening paragraph of Microhydrodynamics].

More than a quarter-century has elapsed since these thoughts in 1987 motivated the writing of a systematic treatise describing the fundamental relationship between the shape of a small, (sub)micron-scale particle and its trajectory in a viscous fluid. Over the subsequent decade, and thanks to generous societal support via the (U.S.) National Science Foundation, this mostly curiosity-driven research program culminated in the definitive analysis of the hydrodynamics of particles with the most extreme shapes: sharp corners and edges. But gradually, curiosity-driven research was overshadowed by projects that arose from the urgency and excitement of translational research: “fluidic self assembly” and “hydrodynamic steering” of proteins, enzymes and even RFID chips.

The unexpected and fortuitous detour in the world of pharmaceutical R&D IT during the genomic revolutionary years of 1997-2003 provided new perspectives including the realization that the field of “computational biology” had devolved into two disjoint communities: one group focused on the “wriggling of molecules” (e.g., molecular dynamics) and the other focused on pattern matching and informatics (e.g., sequence alignment), with the split traceable back to the relative weighting of calculus vs. computer science courses in their respective undergraduate years. But today, thanks to advances in computational technology, a new era can be envisioned that features the reunion of the separated branches of computational biology. We consider some early examples from pharmaceutical informatics of data-intensive informatics approaches applied to “objects” that are not simply numbers or static molecular structures but instead are (computationally-intensive) molecular simulations. These examples hint at new possibilities for rational, computer-aided drug design to help bridge the gap between academic research on discovery biology and search for new molecular entities by the pharmaceutical industry.