NaFM: Pre-training a Foundation Model for Small-Molecule Natural Products

2025-03-22

AI Summary

Existing molecular pretraining models, designed primarily for synthetic compound design, inadequately capture the structural diversity, biosynthetic pathways, and cross-level evolutionary patterns characteristic of natural products (NPs), resulting in poor generalization and suboptimal performance on downstream tasks. To address this, we propose NP-Foundation—the first foundation model specifically tailored for small-molecule NPs. It innovatively integrates skeletal evolutionary priors with side-chain semantic representations and establishes a dual-path pretraining paradigm synergizing contrastive learning and masked graph modeling. Crucially, it is the first to jointly model NP evolution across classification, biosynthetic gene cluster, and microbial host levels. Evaluated on NP classification and target virtual screening, NP-Foundation achieves state-of-the-art performance, significantly outperforming general-purpose molecular models—demonstrating its capacity to deeply encode biosynthetic logic and phylogenetic relationships.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Humans and AI: Other Foundations of Human Computation & AI
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Natural products, as metabolites from microorganisms, animals, or plants, exhibit diverse biological activities, making them crucial for drug discovery. Nowadays, existing deep learning methods for natural products research primarily rely on supervised learning approaches designed for specific downstream tasks. However, such one-model-for-a-task paradigm often lacks generalizability and leaves significant room for performance improvement. Additionally, existing molecular characterization methods are not well-suited for the unique tasks associated with natural products. To address these limitations, we have pre-trained a foundation model for natural products based on their unique properties. Our approach employs a novel pretraining strategy that is especially tailored to natural products. By incorporating contrastive learning and masked graph learning objectives, we emphasize evolutional information from molecular scaffolds while capturing side-chain information. Our framework achieves state-of-the-art (SOTA) results in various downstream tasks related to natural product mining and drug discovery. We first compare taxonomy classification with synthesized molecule-focused baselines to demonstrate that current models are inadequate for understanding natural synthesis. Furthermore, by diving into a fine-grained analysis at both the gene and microbial levels, NaFM demonstrates the ability to capture evolutionary information. Eventually, our method is experimented with virtual screening, illustrating informative natural product representations that can lead to more effective identification of potential drug candidates.

Problem

• Existing deep learning lacks generalizability for natural products.
• Current models fail to capture natural synthesis understanding.
• Molecular characterization methods are unsuitable for natural products.

Innovation

• Pre-trained foundation model for natural products
• Contrastive and masked graph learning objectives
• Captures evolutionary and side-chain information