Welcome to THU-ATOM Lab at the Institute for AI Industry Research (AIR), Tsinghua University.
Led by Professor Yanyan Lan, our lab is at the forefront of AI for science, focusing on pioneering research in biopharmaceuticals and healthcare. We tackle critical challenges in these fields by leveraging advanced data-driven and deep learning methods, offering solutions that surpass traditional approaches in speed, accuracy, and high-throughput capabilities.
Our research spans a broad spectrum of cutting-edge technologies, including representation learning, generative models, deep graph neural networks, and information retrieval applied to molecules and proteins.
Our grand vision is to accelerate drug design, reduce development costs, expand the boundaries of AI-enabled drug research and development, and make significant contributions to human health.
Join us and be part of this revolutionary journey!
- September, 2024 We release the project "Drug The Whole Genome" and introduce our work in paper "Deep contrastive learning enables genome-wide virtual screening".
- August, 2024 Our article titled "Pre-training with Fractional Denoising to Enhance Molecular Property Prediction" has been accepted for publication in Nature Machine Intelligence.
- May, 2024 Our two recent papers have been accepted by ICML 2024, "Rethinking Specificity in SBDD: Leveraging Delta Score and Energy-Guided Diffusion" , and "UniCorn: A Unified Contrastive Learning Approach for Multi-view Molecular Representation Learning" .
- Jan, 2024 Our two recent papers have been accepted by ICLR 2024, "Self-supervised Pocket Pretraining via Protein Fragment-Surroundings Alignment Protein-ligand binding representation learning from fine-grained interactions" , and "Sliced Denoising: A Physics-Informed Molecular Pre-Training Method" .
- Sep, 2023 Our one recent paper have been accepted by NeurIPS 2023, "DrugCLIP: Contrastive Protein-Molecule Representation Learning for Virtual Screening" .
- April, 2023 Our two recent papers have been accepted by ICML 2023, "Fractional Denoising for 3D Molecular Pre-training" , and "Coarse-to-Fine: a Hierarchical Diffusion Model for Molecule Generation in 3D" .