DROIDS

DROIDS (Detecting Relative Outlier Impacts in Dynamic Simulations) analyzes comparative molecular dynamics simulations to detect and quantify residue-level differences in backbone atom bond fluctuations caused by mutations and chemical modifications.

Key Features:

• GPU-accelerated MD support: Uses graphics-processing-unit-accelerated molecular dynamics simulation software, including Amber16, for short-timescale simulations.
• Structural alignment: Employs structural alignment techniques to compare homologous Protein Data Bank structures.
• Statistical testing: Applies Benjamini-Hochberg-corrected Kolmogorov-Smirnov statistics for comparative analyses of dynamic behavior.
• Residue-level resolution: Measures nanosecond-scale atom bond fluctuations along protein backbones at single-amino-acid resolution.
• Mutation and chemical modification analysis: Detects biophysical effects on protein dynamics resulting from single-amino-acid mutations and other chemical modifications.
• Integration with analysis tools: Incorporates cpptraj for trajectory analysis.
• Statistical and visual output: Produces statistical graphics via R graphics and structural visualizations via UCSF Chimera.
• Timescale linkage: Connects short-timescale molecular dynamics processes (measured in femtoseconds) to hypotheses about long-term molecular evolution.

Scientific Applications:

• Molecular evolution: Comparative analysis of dynamic differences between homologous PDB structures to infer evolutionary functional changes.
• Disease-related functional analysis: Investigation of functional changes in proteins caused by genetic mutations and epigenetic modifications.
• Ligand and binding studies: Analysis of binding interactions and dynamic impacts of pharmaceuticals, toxins, and other biomolecules.
• Residue-level mechanistic insight: Identification of specific residues or regions with significant dynamic changes that may underlie functional effects.

Methodology:

Performs structural alignment of homologous PDB structures, runs GPU-accelerated MD simulations (Amber16), analyzes trajectories with cpptraj, applies Benjamini-Hochberg-corrected Kolmogorov-Smirnov tests to nanosecond-scale backbone atom bond fluctuation time series at single-residue resolution, and generates statistical and visual outputs via R graphics and UCSF Chimera.