BigNASim

BigNASim provides a NoSQL database and analysis platform for storage, retrieval, and computational analysis of nucleic acids molecular dynamics (MD) simulation trajectories to support force-field benchmarking, structural analysis, and mechanical-property characterization.

Key Features:

• Database system for nucleic acids simulations: Provides a structured repository tailored to nucleic acids MD trajectories to address gaps in protein-focused trajectory databases.
• Initial dataset and deposition protocol: Hosts an initial benchmarking dataset derived from the parmBSC1 force field comprising 156 simulations totaling over 120 microseconds and supports expansion via a deposition protocol for additional trajectories.
• NoSQL database architecture: Implements a hybrid NoSQL architecture using Cassandra to store raw MD trajectories and MongoDB to store analysis results and simulation metadata.
• Comprehensive analysis tools: Performs analyses including backbone geometries, helical structure parameters, NMR observables, and various mechanical-property calculations.
• Trajectory retrieval and aggregation: Enables retrieval of individual trajectories and assembly of combined meta-trajectories for aggregated analyses.

Scientific Applications:

• Force-field validation and development: Supports validation and development of molecular dynamics force fields such as parmBSC1 by enabling comparison between simulated trajectories and benchmark datasets.
• Comparative structural studies: Enables comparative analyses across different nucleic acid sequences, structures, and simulation conditions using stored trajectories and derived metrics.
• Mechanical and dynamic behavior characterization: Facilitates analysis of mechanical properties and dynamic behaviors of nucleic acids through backbone, helical, NMR, and mechanics analyses.

Methodology:

Uses a hybrid NoSQL storage architecture combining Cassandra for raw MD trajectory storage and MongoDB for analysis results and simulation metadata; performs analyses of backbone geometries, helical structure, NMR observables, and mechanical properties and includes an initial parmBSC1 benchmarking dataset with a deposition protocol for new trajectories.