Clustering algorithms to analyze molecular dynamics simulation trajectories for complex chemical and biological systems

Abstract

Molecular dynamics (MD) simulation has become a powerful tool to investigate the structurefunction relationship of proteins and other biological macromolecules at atomic resolution and biologically relevant timescales. MD simulations often produce massive datasets containing millions of snapshots describing proteins in motion. Therefore, clustering algorithms have been in high demand to be developed and applied to classify these MD snapshots and gain biological insights. There mainly exist two categories of clustering algorithms that aim to group protein conformations into clusters based on the similarity of their shape (geometric clustering) and kinetics (kinetic clustering). In this paper, we review a series of frequently used clustering algorithms applied in MD simulations, including divisive algorithms, agglomerative algorithms (single-linkage, complete-linkage, average-linkage, centroid-linkage and ward-linkage), center-based algorithms (K-Means, K-Medoids, K-Centers, and APM), density-based algorithms (neighbor-based, DBSCAN, density-peaks, and Robust-DB), and spectral-based algorithms (PCCA and PCCA+). In particular, differences between geometric and kinetic clustering metrics will be discussed along with the performances of different clustering algorithms. We note that there does not exist a one-size-fits-all algorithm in the classification of MD datasets. For a specific application, the right choice of clustering algorithm should be based on the purpose of clustering, and the intrinsic properties of the MD conformational ensembles. Therefore, a main focus of our review is to describe the merits and limitations of each clustering algorithm. We expect that this review would be helpful to guide researchers to choose appropriate clustering algorithms for their own MD datasets.

Exploiting a Mechanical Perturbation of a Titin Domain to Identify How Force Field Parameterization Affects Protein Refolding Pathways

“…Analysis algorithms were used to produce the potential of mean force distribution, 55 secondary structure percentage, 56 and clustered structures. 57 Secondary structure percentage was analyzed via DSSP 56 using MDTraj. 58 Clustering of the unfolded structures was done using the KMedoids and KCenters algorithms as implemented in MSMBuilder 59 and MDTraj.…”

“…Two different clustering algorithms were used because the KCenters clustering algorithm commonly picks up outliers in the structures and was thus used to identify the range of structures that were sampled in the unfolded state. 57 Meanwhile, the KMedoids algorithm was used to find the most common unfolded state structures and thus the structures that were most likely to be found in local minima.…”

Exploiting a Mechanical Perturbation of Titin Domain to Identify How Force Field Parameterization Affects Protein Refolding Pathways

“…Analysis algorithms were used to produce the potential of mean force distribution, 55 secondary structure percentage, 56 and clustered structures. 57 Secondary structure percentage was analyzed via DSSP 56 using MDTraj. 58 Clustering of the unfolded structures was done using the KMedoids and KCenters algorithms as implemented in MSMBuilder 59 and MDTraj.…”

Molecular function recognition by supervised projection pursuit machine learning

“…PCA identifies large-scale motions that are assumed critical to function; consequently, functional motions will be misidentified as noise if the dynamics have a smaller amplitudinal variance than what is contained in the top PCA modes 23 . Clustering algorithms are often combined with DR and feature extraction techniques such as PCA in order to identify key conformations that facilitate molecular function 24 , 25 .…”