Optimizing around a 3/3: A research tool for when you become the bottleneck

ORAL

Abstract

The demands on an early career physicists are many. Teaching duties can leave little time left over for research, and even less mental bandwidth. Yet there is an obligation to maintain an active research program for the sake of career advancement. As a visiting professor at a small liberal arts college, this leads to an important question: How do you do research when time, money and attention all scarce? In this contribution, we will present our group's attempt at solving this problem by focusing on developing molecular dynamics simulations that embrace a "hurry up and wait" paradigm, with a focus on efficiently collecting information from slow, long running simulations, and developing tools that make interacting with these simulations more accessible for novice undergraduate researchers. This includes a python program that monitors a running simulation and interacts with it exclusive through existing output files which enables students to interact with legacy code easily (no Fortran required). The same tool also allows us to convert standard molecular dynamics simulations into rare event focused enhanced sampling simulations. The lowered barriers to entry make it easier for students to make productive contributions to ongoing projects, while the rare event monitoring let's us accumulate valuable data to work with once our other demands permit. We offer our approach in the hopes of spurring further conversation about other similar tactics and tools.

Presenters

  • Mitchell Powers

    Gettysburg College

Authors

  • Mitchell Powers

    Gettysburg College

  • Alfredo Roman Jordan

    Gettysburg College