Conditional Latent Space Molecular Scaffold Optimization for Accelerated Molecular Design

2024-11-03

arXiv.org

AI Summary

To address the low generation efficiency and poor practical applicability in drug molecule design, this paper proposes Conditional Latent-Space Bayesian Optimization (C-LSBO): a framework that constructs a conditional variational autoencoder (CVAE) latent space conditioned on atomic environments and performs Bayesian optimization within it to enable minimal, controllable modifications of molecular scaffolds. The method explicitly integrates molecular similarity constraints with objective property optimization (e.g., ADMET), and supports Human-in-the-Loop interaction for real-time expert intervention. Under resource-constrained settings—using lightweight models and limited data—C-LSBO achieves state-of-the-art performance. It generates highly druglike candidates via single-site or substructure-level perturbations, significantly improving both predictive accuracy of target properties and synthetic feasibility.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Constraint Satisfaction and Optimization: Constraint Optimization
• Intelligent Robots: Learning & Optimization for ROB

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions

Abstract

The rapid discovery of new chemical compounds is essential for advancing global health and developing treatments. While generative models show promise in creating novel molecules, challenges remain in ensuring the real-world applicability of these molecules and finding such molecules efficiently. To address this, we introduce Conditional Latent Space Molecular Scaffold Optimization (CLaSMO), which combines a Conditional Variational Autoencoder (CVAE) with Latent Space Bayesian Optimization (LSBO) to modify molecules strategically while maintaining similarity to the original input. Our LSBO setting improves the sample-efficiency of our optimization, and our modification approach helps us to obtain molecules with higher chances of real-world applicability. CLaSMO explores substructures of molecules in a sample-efficient manner by performing BO in the latent space of a CVAE conditioned on the atomic environment of the molecule to be optimized. Our experiments demonstrate that CLaSMO efficiently enhances target properties with minimal substructure modifications, achieving state-of-the-art results with a smaller model and dataset compared to existing methods. We also provide an open-source web application that enables chemical experts to apply CLaSMO in a Human-in-the-Loop setting.

Problem

• Optimizing molecular structures while preserving similarity constraints
• Improving sample efficiency in molecular property optimization
• Enhancing real-world applicability of generated chemical compounds

Innovation

• Conditional VAE integrates with latent space Bayesian optimization
• Latent space optimization enhances molecular sample efficiency
• Preserves molecular similarity while improving target properties