MSMBuilder

MSMBuilder constructs Markov state models and complementary statistical models for high-dimensional time-series data to analyze biomolecular dynamics, including atomistic simulations of protein folding and conformational changes.

Key Features:

• Markov State Models (MSMs): Constructs MSMs to represent stochastic kinetics of biomolecular processes.
• Hidden Markov Models (HMMs): Implements HMMs as a complementary approach for state identification in time-series data.
• Time-structure based Independent Component Analysis (TICA): Applies TICA for time-lagged dimensionality reduction of dynamical data.
• High-dimensional time-series support: Operates on high-dimensional atomistic simulation trajectories and other time-series measurements.
• Atomistic molecular dynamics analysis: Targets analysis of protein folding and conformational changes from MD simulations.
• Interoperability: Designed for integration with existing computational tools used in molecular dynamics studies.

Scientific Applications:

• Protein folding: Derives kinetic models and state ensembles from atomistic folding simulations.
• Conformational change analysis: Identifies and characterizes conformational transitions in biomolecules.
• Molecular dynamics simulation analysis: Extracts kinetic and mechanistic information from MD trajectories.
• General time-series modeling: Applies statistical modeling approaches to computational or experimental high-dimensional time-series beyond biomolecular systems.

Methodology:

Uses construction of Markov state models (MSMs), hidden Markov models (HMMs), and time-structure based independent component analysis (TICA) to model and analyze high-dimensional time-series from atomistic simulations.