deltaGseg: macrostate estimation via molecular dynamics simulations and multiscale time series analysis
Diana H P Low1, Efthymios Motakis
1Institute of Molecular and Cell Biology, Epigenetics, Chromatin and Differentiation, 61 Biopolis Street, Proteos #03-06, Singapore 138673 and Bioinformatics Institute, Genome and Gene Expression Data Analysis, 30 Biopolis Street, Matrix #07-01, Singapore 138671.
Bioinformatics (Oxford, England)
|July 19, 2013
Summary
This study introduces deltaGseg, an R package for analyzing molecular dynamics simulations. It identifies distinct molecular states, offering deeper insights into binding events beyond traditional methods.
Area of Science:
- Computational chemistry
- Biophysics
- Statistical modeling
Background:
- Molecular dynamics simulations are crucial for calculating binding free energies.
- Estimating free energies from multiple simulation series requires robust statistical methods.
- Existing methods often assume a single conformational state, limiting analysis.
Purpose of the Study:
- To develop a more general approach for analyzing binding free energy calculations.
- To identify and quantify multiple statistically distinct subpopulations (macrostates) within simulations.
- To provide tools for molecular biologists and chemists to gain deeper physical insights into molecular interactions.
Main Methods:
- Utilizing statistical modeling for data analysis.
- Applying wavelets denoising for signal processing.
- Employing hierarchical clustering to identify distinct system states.
- Developing the deltaGseg R package for macrostate estimation.
Main Results:
- The deltaGseg R package enables the estimation of macrostates from replicated simulation series.
- The approach accounts for multiple statistically distinct subpopulations, reflecting system macrostates.
- This method offers a more comprehensive analysis compared to traditional moving average techniques.
Conclusions:
- The deltaGseg R package provides a novel approach to analyze binding free energy calculations.
- It allows for the identification of multiple molecular macrostates, enhancing physical insight.
- This tool aids researchers in understanding molecular details not easily accessible through experiments.
