Relaxation predictor - ProkopHapala/FireCore GitHub Wiki

The goal is to be able to quickly extrapolate result of geometry relaxation or trajectory of MD simulation (not after one time-step, but after many time steps).

Physics Simulations with Graph Neural Networks

• Learning Mesh-Based Flow Simulations on Graph Networks, Rayan Kanfar Published in Stanford CS224W GraphML Tutorials
• Learning mesh-based simulation with Graph Networks
  • ArxiV, pdf
  • youtube
• Introduction to basics of Graph Neural Network (GNN)
  • youtube with practical excercise
• Fast software for manipulation with graphs

Differentiable Programing For physical simulations

• DiffTaichi: Differentiable Programming for Physical Simulation (ICLR 2020)
  • arxiv, pdf
  • DiffTaichi:youtube
• Dojo: A Differentiable Physics Engine for Robotics
  • arxiv,pdf
  • youtube

Diferentialbe Projective Dynamics

• github: mit-gfx / diff_pd_public
  • paper DiffPD: Differentiable Projective Dynamics

Physics-informed machine learning

• Physics-informed machine learning, Nature Reviews Physics,3,422–440(2021)
• Physics Informed Machine Learning: High Level Overview of AI and ML in Science and Engineering,youtub,Steve Brunton

Diffusion models for molecular desing and docking

• De novo design of protein structure and function with RFdiffusion

• Diffusion Models in Generative Chemistry for Drug Design

• DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking

• awesome-molecular-generation|

• Graph Diffusion Transformer for Multi-Conditional Molecular Generation

  • github.com/Graph-DiT
  • huggingface/Polymer-Design-With-GraphDiT

AI Chemical Desing

• Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules