Zaixi Zhang

All-Atom Structure Tokenizer: AtomLM-9b for Biomolecular Structure Prediction

The paper introduces AtomLM-9b, a language model using an all-atom structure tokenizer to predict protein structures with performance comparable to AlphaFold 3.

ScienceRSI: Recursive Scientific Self-Improvement

ScienceRSI improves scientific AI through recursive self-improvement by transferring capabilities across model scales, achieving top performance on multiple benchmarks.

Chain of Structures: Connecting Reasoning and Generation for Protein Design

The paper introduces a method for protein design that connects reasoning and generation through structural states, achieving high performance on a benchmark.

Enhancing Pan-cancer Spatial Transcriptomics atSingle-cell Resolution with stPainter

STPAINTER enhances spatial transcriptomics data using a pre-trained scRNA-seq atlas, enabling high-fidelity downstream analysis across cancer types.

BioClaw: Human-Bot Research Collaboration Ecosystems in Group Chats

BioClaw enables human-bot collaboration in group chats to execute biomedical analyses using containerized tools and workflows across multiple domains and data types.

Claw4Science: A Dataset and Platform for the OpenClaw Scientific Agent Ecosystem

The paper introduces a curated dataset and platform for the OpenClaw scientific agent ecosystem, analyzing its structure and challenges in reproducibility and governance.

LabOS: The AI-XR Co-Scientist That Sees and Works With Humans

LabOS is an AI co-scientist that integrates computational reasoning with physical experimentation through multimodal perception and human-AI collaboration, enhancing scientific discovery in biomedical and materials science.

BioLab: End-to-End Autonomous Life Sciences Research with Multi-Agents System Integrating Biological Foundation Models

BioLab uses a multi-agent system with foundation models to autonomously design and validate antibodies, achieving superior performance compared to existing methods.

STELLA: Self-Evolving LLM Agent for Biomedical Research

STELLA is a self-evolving AI agent that improves through interaction, achieving high performance in biomedical tasks and driving experimental discoveries in oncology and protein engineering.

BioMiner: A Multi-modal System for Automated Mining of Protein-Ligand Bioactivity Data from Literature

BioMiner automates extraction of protein-ligand bioactivity data from literature, achieving an F1 score of 0.32 and improving downstream model performance by 3.9%.

Learning Universal Representations of Intermolecular Interactions with ATOMICA

ATOMICA learns transferable representations of intermolecular interactions across diverse molecular types, enabling cross-modal comparisons and experimentally testable hypotheses.

RNAGenesis: A Generalist Foundation Model for Functional RNA Therapeutics

RNAGenesis is a foundation model for RNA therapeutics that integrates sequence, structure, and function prediction, achieving state-of-the-art performance on multiple tasks and enabling functional RNA design.

Efficient Generation of Protein Pockets with PocketGen

PocketGen generates high-fidelity protein pockets with superior binding affinity and structural validity, operating ten times faster than physics-based methods.

FoldMark: Safeguarding Protein Structure Generative Models with Distributional and Evolutionary Watermarking

This paper introduces FoldMark, a watermarking method for protein generative models that ensures biosecurity by enabling user tracing and maintaining structural quality.

An Equivariant Generative Framework for Molecular Graph-Structure Co-Design

This paper introduces MolCode, a generative framework for molecule design that integrates 2D graph and 3D geometry information, achieving high validity and affinity in generated molecules.