Tios: The Internet of Simulations. Turning Molecular Dynamics into a Data Streaming Web Application

Area of Science:

• Computational chemistry
• Biophysics
• Scientific computing

Background:

• Current high-performance computing (HPC) configurations often impose suboptimal workflows for molecular dynamics (MD) simulations.
• Traditional batch job systems and post-processing analysis hinder real-time data collection and adaptive sampling for MD simulations.

Purpose of the Study:

• To develop a novel application, Tios, that enables real-time data streaming from MD simulations.
• To provide researchers with flexible, remote access to ongoing simulations for improved data analysis and sampling strategies.

Main Methods:

• Tios is a Python application designed to convert MD simulations into discoverable, streaming web applications.
• It integrates with standard MD codes without requiring modifications, facilitating adoption.
• The application leverages Internet of Things principles for remote accessibility and real-time data retrieval.

Main Results:

• Tios successfully transforms MD simulations into accessible web services.
• Demonstrated capability to stream simulation data in real time, allowing researchers to connect and download data as needed.
• Test cases confirm the application's effectiveness and ease of use with standard MD codes.

Conclusions:

• Tios offers a flexible and efficient alternative to traditional HPC-based MD simulation workflows.
• The application enhances the ability to generate representative Boltzmann-weighted ensembles by enabling real-time data analysis and adaptive simulation termination.
• Tios promotes a more dynamic and interactive approach to molecular dynamics research.