Tcmscience builds autonomous agent for reliable traditional medicine research
TCMScience: An Autonomous Scientist for Traditional Chinese Medicine
Machine Learning
Summary
Researching Traditional Chinese Medicine involves many types of evidence, like ancient texts, experiments, and clinical data. The authors created TCMScience, a computer agent that carefully checks and combines these different sources while making sure claims are backed by solid evidence. It uses a system that separates suggestions from what is actually done and only shares findings that meet strict rules. This method helps avoid mistakes and ensures conclusions are trustworthy. They demonstrated this with detailed tests and a large collection of data and tools.
What this means in practice
- •For pharmaceutical data teams: Develop automated systems that verify and integrate diverse biomedical data for traditional medicine drug discovery with controlled evidence verification.
- •For biomedical software engineers: Build modular platforms allowing complex scientific workflows that enforce provenance and claim checks separating proposal from execution authority.
Authors
Yanlan Kang, Ruiqi Liu, Xinpei Gu, Tzu-Yi Lei, Shuai Xu, Xukun Zhang, William Cheng-Chung Chu
Abstract
Traditional Chinese medicine (TCM) research connects classical texts, medicinal entities, molecular assays, computational predictions, and clinical studies. An autonomous research agent must preserve the inferential boundaries between these evidence types while coordinating analysis across heterogeneous resources. We introduce TCMScience, a scientific knowledge-discovery agent that separates language-model proposals from execution and release authority. Its architecture combines a trusted kernel, a scientific compiler, and claim-support contracts: plans are compiled into typed operations, execution is bounded by explicit permissions, and candidate findings are checked against source provenance, study design, and claim scope before release. The full capability registry contains 117 public-source connectors with 444 typed operations across 29 domains and 150 native tools across 12 domains; the TCM data hub catalogs 136 databases, 76 dataset specifications, and 84,294 formula-source records. These resources support model-independent workflows through browser and local execution interfaces. In a network-pharmacology analysis of Gegen Qinlian Tang, replacing the whole-proteome background with the assayed-protein background changed the number of significant pathways from 66 to zero; incorporating inactive screening records identified nine pathways, including drug-metabolism signals retained as mechanistic hypotheses. A governance ablation over 121 constructed erroneous outputs spanning 26 classes released 121 errors without governance, 57 with the domain layer, seven with the kernel, and zero with both layers. All 10 compliant controls were retained. These results establish complementary control of known failure modes and show how explicit evidence contracts change the conclusions an agent is permitted to report.