MR-MAS: an end-to-end multi-agent system for autonomous Mendelian randomization analysis.
6 affiliations
- 1College of Informatics, Huazhong Agricultural University, Wuhan, 430070, China.
- 2College of Plant Science & Technology, Huazhong Agricultural University, Wuhan, 430070, China.
- 3College of Engineering and Information Technology, Adelaide University, Adelaide, SA 5095, Australia.
- 4Business Banking, Commonwealth Bank, Sydney, NSW 2000, Australia.
- 5School of Computer Science and Technology, Hainan University, Haikou, 570228, China.
- 6Hubei Key Laboratory of Agricultural Bioinformatics, Huazhong Agricultural University, Wuhan, 430070, China.
Abstract
Motivation
Understanding disease-related causal mechanisms is essential for biomedical research, yet the practical use of Mendelian Randomization (MR) remains constrained by fragmented analytical workflows and a strong dependence on specialized expertise. These challenges limit the efficient, standardized, and reproducible execution of end-to-end MR analyses.Results
We propose MR-MAS, an end-to-end automated MR analysis system built on a multi-agent architecture. MR-MAS leverages the reasoning and coordination capabilities of large language models (LLMs) and integrates external tools to unify data acquisition, causal effect estimation, results interpretation, and manuscript preparation into a single workflow, thereby reducing manual effort. MR-MAS further incorporates a structured Mendelian Randomization Knowledge Base (MRKB) to support MR method implementation and targeted repair of MR analysis code. Across diverse test scenarios, MR-MAS demonstrated stable workflow execution, reproduced several literature-supported causal signals, and maintained consistent causal estimates across repeated runs and different model backends. Additional evaluations showed robust trait retrieval and improved code-repair performance with MRKB guidance. These results suggest that MR-MAS may provide practical support for scalable, standardized, and reproducible MR analyses in biomedical research.Availability and implementation
The source code, MRKB, and benchmark data are freely available at https://github.com/Zaiwen/MR-MAS, with the version used in this study archived at https://doi.org/10.5281/zenodo.20342636.
Related
- Hypoglycemic Agent Targets and Susceptibility to Lower Respiratory Infection: A Pharmacogenomic Mendelian Randomization Study
- Monocyte Immunometabolic Axis Linking Type 2 Diabetes to Coronary Artery Disease Revealed by Multi-Omics Integration and Mendelian Randomization
- Multi-cohort meta-analysis, single-cell and Mendelian randomization analyses with experimental validation identify RHBDF2 as a shared mediator of ulcerative colitis and blood-brain barrier disruption.
- Integrated Single-Cell Transcriptomics, Multi-Omics, and Mendelian Randomization Reveal Ferroptosis-Related Molecular Subtypes and a Genetically Supported Causal Candidate in Psoriasis.
- Associations of Blood Lipid Traits with Cognitive-Related Outcomes and Their Sensitivity to APOE-Region Instrumental Variables: A Mendelian Randomization and Colocalization Study
BibTeX
@article{wei2026mrmas,
title = {{MR-MAS: an end-to-end multi-agent system for autonomous Mendelian randomization analysis.}},
author = {Wendi Wei and Yuling Fan and Lin Li and Thuc Duy Le and Xiaomei Li and Debo Cheng and Zaiwen Feng},
journal = {Bioinformatics (Oxford, England)},
year = {2026},
month = oct,
doi = {10.1093/bioinformatics/btag722},
url = {https://europepmc.org/article/MED/42895113},
note = {CC BY},
}
RIS
TY - JOUR
TI - MR-MAS: an end-to-end multi-agent system for autonomous Mendelian randomization analysis.
AU - Wendi Wei
AU - Yuling Fan
AU - Lin Li
AU - Thuc Duy Le
AU - Xiaomei Li
AU - Debo Cheng
AU - Zaiwen Feng
PY - 2026
DA - 2026/10/01
JO - Bioinformatics (Oxford, England)
DO - 10.1093/bioinformatics/btag722
UR - https://europepmc.org/article/MED/42895113
AB - Motivation Understanding disease-related causal mechanisms is essential for biomedical research, yet the practical use of Mendelian Randomization (MR) remains constrained by fragmented analytical workflows and a strong dependence on specialized expertise. These challenges limit the efficient, standardized, and reproducible execution of end-to-end MR analyses.Results We propose MR-MAS, an end-to-end automated MR analysis system built on a multi-agent architecture. MR-MAS leverages the reasoning and coordination capabilities of large language models (LLMs) and integrates external tools to unify data acquisition, causal effect estimation, results interpretation, and manuscript preparation into a single workflow, thereby reducing manual effort. MR-MAS further incorporates a structured Mendelian Randomization Knowledge Base (MRKB) to support MR method implementation and targeted repair of MR analysis code. Across diverse test scenarios, MR-MAS demonstrated stable workflow execution, reproduced several literature-supported causal signals, and maintained consistent causal estimates across repeated runs and different model backends. Additional evaluations showed robust trait retrieval and improved code-repair performance with MRKB guidance. These results suggest that MR-MAS may provide practical support for scalable, standardized, and reproducible MR analyses in biomedical research.Availability and implementation The source code, MRKB, and benchmark data are freely available at https://github.com/Zaiwen/MR-MAS, with the version used in this study archived at https://doi.org/10.5281/zenodo.20342636.
SN - 1367-4803
SN - 1367-4811
ER -