Metric-Based Assessment Method for MS-T Formalism with Small Subsets of Torsional Conformers.

J Phys Chem A

Department of Mechanical Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon999077, Hong Kong.

Published: November 2022

The multi-structural approximation with torsional anharmonicity (MS-T) method and its variants have been widely used for calculating conformational-rovibrational partition functions of large molecules. The present work aimed to propose a systematic method to assess and explain the performance of various variants of the MS-T method. First, we proposed the simplest variant MS-T(2NN) (two nearest neighborhood torsions are coupled) and systematically validated it for large alkanes -CH ( = 6-10) and their transition states of hydrogen abstraction reactions. Second, we proposed a metric-based method to explain the underlying reason for the good performance of MS-T(2NN)─it includes the torsional conformers that have dominant contributions to the partition function calculations. These conformers are closer to the lowest-energy conformer in the space of dihedral and energy metrics. Third, the same observation and explanation apply to the other two variants, MS-2DT (any two torsions are coupled) and MS-3DT (any three torsional are coupled), which contain increasingly more torsional conformers than MS-T(2NN) but are subsets of the complete set of torsional conformers considered by the MS-T method. Overall, the present method provides a mathematically rigorous and computationally effective diagnosis tool to assess various MS-T methods dealing with the torsional anharmonicity of large molecules in the partition function calculation.

Download full-text PDF

Source
http://dx.doi.org/10.1021/acs.jpca.2c04724DOI Listing

Publication Analysis

Top Keywords

torsional conformers
16
ms-t method
12
torsional anharmonicity
8
large molecules
8
torsions coupled
8
partition function
8
method
7
torsional
7
ms-t
5
conformers
5

Similar Publications

A force field is a critical component in molecular dynamics simulations for computational drug discovery. It must achieve high accuracy within the constraints of molecular mechanics' (MM) limited functional forms, which offers high computational efficiency. With the rapid expansion of synthetically accessible chemical space, traditional look-up table approaches face significant challenges.

View Article and Find Full Text PDF

FlowPacker: Protein side-chain packing with torsional flow matching.

Bioinformatics

January 2025

Department of Molecular Genetics, University of Toronto, Ontario, M5S 3K3, Canada.

Motivation: Accurate prediction of protein side-chain conformations is necessary to understand protein folding, protein-protein interactions and facilitate de novo protein design.

Results: Here we apply torsional flow matching and equivariant graph attention to develop FlowPacker, a fast and performant model to predict protein side-chain conformations conditioned on the protein sequence and backbone. We show that FlowPacker outperforms previous state-of-the-art baselines across most metrics with improved runtime.

View Article and Find Full Text PDF
Article Synopsis
  • GADIFF is a novel graph attention diffusion model designed for generating molecular conformations, leveraging equivariant networks and self-attention mechanisms to enhance feature representation and noise prediction.
  • The model incorporates Graph Isomorphism Networks for capturing local interactions within molecular structures and demonstrates improved performance in generating diverse and accurate molecular conformations compared to existing state-of-the-art methods.
  • A derived model, GADIFF-NCI, extends GADIFF's capabilities to noncovalent interaction systems, showing effective conformation generation, indicating its potential for broader applications in studying complex molecular conformations.
View Article and Find Full Text PDF

Decoding KRAS dynamics: Exploring the impact of mutations and inhibitor binding.

Arch Biochem Biophys

December 2024

Department of Pharmaceutical Sciences, School of Health Sciences and Technology, UPES, Dehradun, 248007, Uttarakhand, India. Electronic address:

KRAS (Kirsten rat sarcoma viral oncogene homologue), the most common mutated protein in human cancers, is the leading cause of morbidity and mortality. Before Sotorasib (AMG-510) was approved for non-small cell lung cancer treatment in 2020, the oncogenic KRAS mutations were believed to be non-druggable. High-resolution X-ray crystal structures of GDP-bound KRAS mutants with and without inhibitor are resolved and deposited in the Protein Data Bank (PDB).

View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!