The primary goal of this collaborative project, led by the team at The University of Chicago, was to create a parallel scripting language, called Parsl, that could be widely used across many different scientific disciplines for high performance computing. The project aimed to meet the needs of a wide variety of users across many disciplines, and to provide sample workflows, documentation, and other training materials that would aid in the adoption of this parallel scripting platform.

The team at The College of New Jersey (TCNJ) was one of several collaborating institutions that worked alongside chemistry teams from Colorado State University, Northern Arizona University (NAU), and California State University at Los Angeles to develop computational chemistry workflows that could be translated into the Parsl language. TCNJ is a residential, primarily undergraduate institution with students from diverse backgrounds, and the Chemistry department at TCNJ provides American Chemical Society (ACS) certified degrees to a large number of majors each year, in additional to serving several hundred TCNJ students from other departments. Therefore, the project also had a significant impact on the training of undergraduate students at TCNJ in PI Baker?s computational chemistry laboratory over the course of the project period.

The computational chemistry research enabled by this project involved studying the interactions of proteins and lipids in novel liquid environments. More specifically, the TCNJ and NAU teams have worked together to understand how liquids composed of charged molecules (e.g., ionic liquids and/or deep eutectic solvents) can influence protein structure and membrane assembly and permeability. From our studies of model cell membranes in a deep eutectic solvent we discovered that the presence of the solvent impacted important membrane physical properties, and in our study of an intrinsically disordered protein in an ionic liquid we discovered that the solvent could shift the types of structures that the protein would adopt compared to being in water. The computational chemistry workflow for these types of simulations has been ported into Parsl, which allows for a more streamlined setup and execution of such simulations on high performance computing systems. This research has also led to additional directions exploring how solvent-modified membrane assembly, and how other proteins such as insulin have their structure and aggregation impacted in the specialized solvents. Additional publications related to these directions are also anticipated in the future, and will include undergraduate student authors.

The impact of this project on the development of undergraduate students at TCNJ has been significant. At TCNJ eight undergraduates have worked directly on the ionic liquid and deep eutectic solvent projects. Among them they included students from underrepresented groups in the sciences, first-generation college students, and students with backgrounds in diverse majors from physics, to chemistry, to biology. Students learned how to use high performance computing, how to set up and carry out computational chemistry calculations, and how to write scripts in Python, amongst other skills. Students also benefited from interacting with collaborators at other project sites. Undergraduates working on this project have gone on to pursue PhDs in fields such as chemistry and materials science, medical school, and dental school, and have earned awards such as the ACS Priscilla Carney Jones award, the NSF GRF award, and NSF GRF and Goldwater honorable mentions. Students have also presented their research at local, regional, and national conferences, and have been included on publications.

Last Modified: 01/08/2021
Modified by: Joseph Baker