Selecting Features for Markov Modeling: A Case Study on HP35

TL;DR

This paper investigates how to select appropriate features for Markov state models to accurately represent the folding dynamics of HP35, emphasizing the importance of feature choice in capturing the energy landscape and timescales.

Contribution

It demonstrates the effectiveness of different feature types, like dihedral angles and tertiary contacts, in constructing mechanistically meaningful Markov models for protein folding.

Findings

Dihedral angles effectively capture native energy basin.

Tertiary contacts better describe unfolded states and folding process.

Feature selection influences the accuracy of timescale reproduction.

Abstract

Markov state models represent a popular means to interpret molecular dynamics trajectories in terms of memoryless transitions between metastable conformational states. To provide a mechanistic understanding of the considered biomolecular process, these states should reflect structurally distinct conformations and ensure a timescale separation between fast intrastate and slow interstate dynamics. Adopting the folding of villin headpiece (HP35) as a well-established model problem, here we discuss the selection of suitable input coordinates or `features', such as backbone dihedral angles and interresidue distances. We show that dihedral angles account accurately for the structure of the native energy basin of HP35, while the unfolded region of the free energy landscape and the folding process are best described by tertiary contacts of the protein. To construct a contact-based model, we consider various ways to define and select contact distances, and introduce a low-pass filtering of the feature trajectory as well as a correlation-based characterization of states. Relying on input data that faithfully account for the mechanistic origin of the studied process, the states of the resulting Markov model are clearly discriminated by the features, describe consistently the hierarchical structure of the free energy landscape, and$\unicode{0x2014}$as a consequence$\unicode{0x2014}$correctly reproduce the slow timescales of the process.