MolTextNet: A Two-Million Molecule-Text Dataset for Multimodal Molecular Learning
By: Yihan Zhu , Gang Liu , Eric Inae and more
Potential Business Impact:
Helps find new medicines by understanding molecule descriptions.
Small molecules are essential to drug discovery, and graph-language models hold promise for learning molecular properties and functions from text. However, existing molecule-text datasets are limited in scale and informativeness, restricting the training of generalizable multimodal models. We present MolTextNet, a dataset of 2.5 million high-quality molecule-text pairs designed to overcome these limitations. To construct it, we propose a synthetic text generation pipeline that integrates structural features, computed properties, bioactivity data, and synthetic complexity. Using GPT-4o-mini, we create structured descriptions for 2.5 million molecules from ChEMBL35, with text over 10 times longer than prior datasets. MolTextNet supports diverse downstream tasks, including property prediction and structure retrieval. Pretraining CLIP-style models with Graph Neural Networks and ModernBERT on MolTextNet yields improved performance, highlighting its potential for advancing foundational multimodal modeling in molecular science. Our dataset is available at https://huggingface.co/datasets/liuganghuggingface/moltextnet.
Similar Papers
BioMedGPT-Mol: Multi-task Learning for Molecular Understanding and Generation
Artificial Intelligence
Teaches computers to invent new medicines.
NovoMolGen: Rethinking Molecular Language Model Pretraining
Machine Learning (CS)
Creates new medicines faster and better.
NovoMolGen: Rethinking Molecular Language Model Pretraining
Machine Learning (CS)
Finds new medicines faster by reading molecule language.