Learning Disentangled Equivariant Representation for Explicitly Controllable 3D Molecule Generation
This addresses the need for controllable molecule generation in drug discovery, representing an incremental advance by combining existing techniques like disentanglement and equivariance.
The paper tackles the problem of generating 3D drug-like molecules with explicit control over properties like druglikeness and binding affinity, achieving validated effectiveness in property-guided and context-guided generation for de-novo design and drug discovery.
We consider the conditional generation of 3D drug-like molecules with \textit{explicit control} over molecular properties such as drug-like properties (e.g., Quantitative Estimate of Druglikeness or Synthetic Accessibility score) and effectively binding to specific protein sites. To tackle this problem, we propose an E(3)-equivariant Wasserstein autoencoder and factorize the latent space of our generative model into two disentangled aspects: molecular properties and the remaining structural context of 3D molecules. Our model ensures explicit control over these molecular attributes while maintaining equivariance of coordinate representation and invariance of data likelihood. Furthermore, we introduce a novel alignment-based coordinate loss to adapt equivariant networks for auto-regressive de-novo 3D molecule generation from scratch. Extensive experiments validate our model's effectiveness on property-guided and context-guided molecule generation, both for de-novo 3D molecule design and structure-based drug discovery against protein targets.