Integrating D-MPNN and molecular dynamics for chemical interaction prediction

Overview

This innovative solution leverages a hybrid approach combining Directed Message Passing Neural Networks (D-MPNN) and molecular dynamics (MD) simulations to accurately predict the interactions of chemical compounds with various substrates, such as hair, skin, and textiles. By integrating these techniques, it facilitates the discovery of structurally novel and functionally superior compounds. This method is particularly transformative for designing sustainable and innovative cosmetic and detergent formulations.

Technical specifications

Key features:

• D-MPNN Integration: Utilizes graph-based data for learning molecular representations, enhancing the prediction of molecular properties.
• Molecular Dynamics Simulations: Provides atomic-level insights into adsorption, diffusion, and conformational changes, crucial for understanding chemical interactions.
• QSAR Analysis: Aids in predicting key functionalities like film formation, color retention, and viscosity changes.
• Model Development: Involves building a D-MPNN framework, conducting virtual screenings, and performing MD simulations using GROMACS and Gaussian tools.

Technology readiness level

Currently, the technology is at TRL 2, indicating that the concept has been formulated and initial validation processes are underway. Future steps include experimental validation and chemical synthesis to further advance the technology readiness.