EUVIMEDEUVIMED entdecken
Crossref

An exploration of machine learning models for the determination of reaction coordinates associated with conformational transitions

Nawavi Naleem iD, Charlles R A Abreu iD, Krzysztof Warmuz iD, Muchen Tong iD, Serdal Kirmizialtin iD, Mark E Tuckerman iD

The Journal of Chemical Physics · 2023

DOI: 10.1063/5.0147597 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The Journal of Chemical Physics, 2023. Verfasst von Nawavi Naleem und 5 weiteren Autor:innen. Beteiligte Einrichtung: Chemistry Program, Science Division, New York University, Abu Dhabi, UAE. Hauptthemen laut MEDLINE: Sarcosine, Dipeptides. Die Arbeit wurde bisher 23-mal zitiert (Crossref) und nennt 112 Literaturangaben. Gefördert von: National Science Foundation, New York University Abu Dhabi, U.S. Department of Energy.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Determining collective variables (CVs) for conformational transitions is crucial to understanding their dynamics and targeting them in enhanced sampling simulations. Often, CVs are proposed based on intuition or prior knowledge of a system. However, the problem of systematically determining a proper reaction coordinate (RC) for a specific process in terms of a set of putative CVs can be achieved using committor analysis (CA). Identifying essential degrees of freedom that govern such transitions using CA remains elusive because of the high dimensionality of the conformational space. Various schemes exist to leverage the power of machine learning (ML) to extract an RC from CA. Here, we extend these studies and compare the ability of 17 different ML schemes to identify accurate RCs associated with conformational transitions. We tested these methods on an alanine dipeptide in vacuum and on a sarcosine dipeptoid in an implicit solvent. Our comparison revealed that the light gradient boosting machine method outperforms other methods. In order to extract key features from the models, we employed Shapley Additive exPlanations analysis and compared its interpretation with the “feature importance” approach. For the alanine dipeptide, our methodology identifies ϕ and θ dihedrals as essential degrees of freedom in the C7ax to C7eq transition. For the sarcosine dipeptoid system, the dihedrals ψ and ω are the most important for the cisαD to transαD transition. We further argue that analysis of the full dynamical pathway, and not just endpoint states, is essential for identifying key degrees of freedom governing transitions.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

SarcosineDipeptidesMolecular ConformationSolventsAlanine

Substanzen

SarcosineDipeptidesSolventsAlanine

Autor:innen und Institutionen

  • Chemistry Program, Science Division, New York University, Abu Dhabi, UAE.
  • Chemical Engineering Department, Escola de Química, Universidade Federal do Rio de Janeiro, 21941-909 Rio de Janeiro, RJ, Brazil.
  • Computer Science Program, Science Division, New York University, Abu Dhabi, UAE.
  • Department of Chemistry, New York University (NYU), New York, New York 10003, USA.
  • Center for Smart Engineering Materials, New York University, Abu Dhabi, UAE.
  • Courant Institute of Mathematical Sciences, New York University, New York, New York 10012, USA.

ORCID-Profile: Nawavi Naleem, Charlles R A Abreu, Krzysztof Warmuz, Muchen Tong, Serdal Kirmizialtin, Mark E Tuckerman

Förderung

  • National Science Foundation
  • New York University Abu Dhabi
  • U.S. Department of Energy

Zitieren

APA

Naleem, N., Abreu, C. R. A., Warmuz, K., Tong, M., Kirmizialtin, S., & Tuckerman, M. E. (2023). An exploration of machine learning models for the determination of reaction coordinates associated with conformational transitions. The Journal of Chemical Physics. https://doi.org/10.1063/5.0147597

Vancouver

Naleem N, Abreu CRA, Warmuz K, Tong M, Kirmizialtin S, Tuckerman ME. An exploration of machine learning models for the determination of reaction coordinates associated with conformational transitions. The Journal of Chemical Physics. 2023. doi:10.1063/5.0147597

BibTeX
@article{Naleem2023,
  author = {Naleem, Nawavi and Abreu, Charlles R A and Warmuz, Krzysztof and Tong, Muchen and Kirmizialtin, Serdal and Tuckerman, Mark E},
  title = {An exploration of machine learning models for the determination of reaction coordinates associated with conformational transitions},
  journal = {The Journal of Chemical Physics},
  year = {2023},
  doi = {10.1063/5.0147597}
}
Kostenlos starten →