Drug Discovery SMILES-to-Pharmacokinetics Diffusion Models with Deep Molecular Understanding

2024-08-14

arXiv.org

AI Summary

Pharmacokinetic (PK) data across drug datasets exhibit sparse overlap, hindering multi-drug combination studies and high-throughput screening. To address this, we propose Imagand—the first SMILES-conditioned diffusion generative model for PK properties—integrating syntax-aware SMILES encoding with joint multi-property modeling to enable end-to-end, structure–property controllable generation. Our framework effectively mitigates the sparsity of real-world PK data distributions, achieving high fidelity in both univariate and bivariate property distributions. Generated synthetic data significantly enhance downstream PK prediction performance: MAE reductions of 12.6–28.3% are observed for clearance (CL), volume of distribution at steady state (VDss), and half-life (T₁/₂). The implementation is publicly available and has been adopted by the research community.

Technology Category

• Natural Language Processing: Generation
• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Web Mining and Content Analysis: Web data generation and simulation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Artificial intelligence (AI) is increasingly used in every stage of drug development. One challenge facing drug discovery AI is that drug pharmacokinetic (PK) datasets are often collected independently from each other, often with limited overlap, creating data overlap sparsity. Data sparsity makes data curation difficult for researchers looking to answer research questions in poly-pharmacy, drug combination research, and high-throughput screening. We propose Imagand, a novel SMILES-to-Pharmacokinetic (S2PK) diffusion model capable of generating an array of PK target properties conditioned on SMILES inputs. We show that Imagand-generated synthetic PK data closely resembles real data univariate and bivariate distributions, and improves performance for downstream tasks. Imagand is a promising solution for data overlap sparsity and allows researchers to efficiently generate ligand PK data for drug discovery research. Code is available at url{https://github.com/bing1100/Imagand}.

Problem

• Addresses data sparsity in drug pharmacokinetic datasets
• Generates synthetic PK properties from SMILES inputs
• Improves downstream drug discovery research efficiency

Innovation

• SMILES-to-PK diffusion model for drug discovery
• Generates synthetic PK data resembling real distributions
• Improves downstream task performance with sparse data