MING: A Functional Approach to Learning Molecular Generative Models

2024-10-16

arXiv.org

AI Summary

Traditional molecular generation methods are constrained by the limited expressiveness of sequence- or graph-based representations and difficulties in enforcing equivariance. This paper introduces the first function-representation-based paradigm for molecular generation, modeling molecules as input-output mapping functions and constructing a diffusion model directly in function space—jointly denoising both domain and codomain. Our approach integrates implicit neural representations (INRs) for continuous, function-level modeling and proposes an EM-optimized joint denoising mechanism that balances modeling fidelity and architectural simplicity. On multiple benchmark datasets, our method surpasses state-of-the-art data-space approaches: it significantly improves chemical validity of generated molecules, accelerates inference by 2.1–3.8×, and reduces parameter count by 37%–59%. These results empirically validate the effectiveness and efficiency of function-space modeling for molecular generation.

Technology Category

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

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Web Mining and Content Analysis: Web data generation and simulation
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Traditional molecule generation methods often rely on sequence- or graph-based representations, which can limit their expressive power or require complex permutation-equivariant architectures. This paper introduces a novel paradigm for learning molecule generative models based on functional representations. Specifically, we propose Molecular Implicit Neural Generation (MING), a diffusion-based model that learns molecular distributions in the function space. Unlike standard diffusion processes in the data space, MING employs a novel functional denoising probabilistic process, which jointly denoises information in both the function's input and output spaces by leveraging an expectation-maximization procedure for latent implicit neural representations of data. This approach enables a simple yet effective model design that accurately captures underlying function distributions. Experimental results on molecule-related datasets demonstrate MING's superior performance and ability to generate plausible molecular samples, surpassing state-of-the-art data-space methods while offering a more streamlined architecture and significantly faster generation times. The code is available at https://github.com/v18nguye/MING.

Problem

• Introduces functional molecular generation approach
• Proposes diffusion-based model in function space
• Enables accurate, efficient molecular sample generation

Innovation

• Functional molecular representations
• Diffusion-based functional denoising
• Implicit neural latent modeling