Skip to yearly menu bar Skip to main content


Poster
in
Workshop: AI for Accelerated Materials Design (AI4Mat-2023)

Towards equilibrium molecular conformation generation with GFlowNets

Alexandra Volokhova · MichaƂ Koziarski · Alex Hernandez-Garcia · Chenghao Liu · Santiago Miret · Pablo Lemos · Luca Thiede · Zichao Yan · Alan Aspuru-Guzik · Yoshua Bengio

Keywords: [ molecular conformation generation ] [ gflownet ] [ Boltzmann generator ] [ GFlowNet ]


Abstract:

Sampling diverse, thermodynamically feasible molecular conformations plays a crucial role in predicting properties of a molecule. In this paper we propose to use GFlowNet for sampling conformations of small molecules from the Boltzmann distribution, as determined by the molecule's energy. The proposed approach can be used in combination with energy estimation methods of different fidelity and discovers a diverse set of low-energy conformations for highly flexible drug-like molecules. We demonstrate that GFlowNet can reproduce molecular potential energy surfaces by sampling proportionally to the Boltzmann distribution.

Chat is not available.