Naturally occurring enzymes are widely used as practical, environmentally friendly alternatives to traditional catalysts. However, destabilization of enzymes in nonbiological environments limits their utility. Synthetic polymers have proven effective to stabilize enzymes under environmental stress, although details of the mechanism of stabilization are still unclear, posing a significant challenge for the rational design of polymers for enzyme encapsulation. We have recently designed a combined experimental and computational approach based on an active learning feedback loop scheme that informs rational polymer synthesis for single-enzyme encapsulation. As part of that effort, here we present an algorithm that automates the generation of quantum-mechanical data from multiple conformers of monomers aimed at enriching the description of the chemical features of synthetic polymer candidates. The algorithm is implemented as an open-source Python 3 program, Smiles4Psi, that uses some of the functionalities of the Open Babel library. Smiles4Psi is initially interfaced with the quantum chemistry package Psi4, and the computations required to populate the database are independent of each other, leading to a naturally parallel approach. We have exploited this advantage and developed an Azure cloud computing framework to generate a quantum-mechanical information database for methacrylate-based polymers at the ωB97X-D/aug-cc-pVTZ level of theory. We use this database as a reference set to study alternative levels of theory that may be suitable for quantum-mechanical polymer featurization.

Smiles4Psi: Automated conformational search with a genetic algorithm interfaced with a Psi4 workflow manager quantum-mechanical featurization of synthetic polymers

ACS Fall 2021

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.