---
title: "TxAgent — paper by Gao et al. (2025) — summary and ideas"
description: "The authors present TxAgent, an agent built on a fine-tuned Llama-3.1-8B-Instruct. 7.5 points. 24 references, cited by 7 papers here."
authors: ["Shanghua Gao", "Richard Zhu", "Zhenglun Kong", "Ayush Noori", "Xiaorui Su", "Curtis Ginder", "Theodoros Tsiligkaridis", "Marinka Zitnik"]
year: 2025
venue: "arXiv:2503.10970"
source: "https://arxiv.org/abs/2503.10970"
doi: "10.48550/arxiv.2503.10970"
arxiv: "2503.10970"
licence: "CC BY 4.0 (arXiv)"
licence_url: "http://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/gao-2025-txagent-ai-agent-therapeutic-reasoning/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/gao-2025-txagent-ai-agent-therapeutic-reasoning/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2025](https://md.research.elamur.ai/free/en/papers/2025/page/7/) / TxAgent: An AI Agent for Therapeutic Reasoning Across a Universe of Tools

# TxAgent: An AI Agent for Therapeutic Reasoning Across a Universe of Tools

**Authors:** [Shanghua Gao](https://md.research.elamur.ai/free/en/authors/shanghua-gao/), [Richard Zhu](https://md.research.elamur.ai/free/en/authors/richard-zhu/), Zhenglun Kong, [Ayush Noori](https://md.research.elamur.ai/free/en/authors/ayush-noori/), Xiaorui Su, Curtis Ginder, Theodoros Tsiligkaridis and [Marinka Zitnik](https://md.research.elamur.ai/free/en/authors/marinka-zitnik/) · **Year:** 2025 · **Venue:** arXiv:2503.10970 · **Licence:** [CC BY 4.0 (arXiv)](http://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2503.10970) · [PDF](https://arxiv.org/pdf/2503.10970) · [DOI 10.48550/arxiv.2503.10970](https://doi.org/10.48550/arxiv.2503.10970) · [arXiv 2503.10970](https://arxiv.org/abs/2503.10970) · [OpenAlex](https://openalex.org/W4417060272)

**Citations:** 10 citations (OpenAlex, 2026-10-14) · [7](https://md.research.elamur.ai/free/en/papers/gao-2025-txagent-ai-agent-therapeutic-reasoning/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
Picking the right drug for a particular patient means weighing comorbidities, co-medications, interactions and contraindications. General-purpose LLMs handle this poorly: they hallucinate, lack up-to-date knowledge, and struggle to combine many clinical variables. The authors present TxAgent, an agent built on a fine-tuned Llama-3.1-8B-Instruct. It interleaves explicit reasoning steps with calls to external biomedical resources and returns a transparent trace alongside its answer. Its tools come from ToolUniverse, a curated set drawing on openFDA (all FDA-approved drugs since 1939), Open Targets and the Human Phenotype Ontology. A ToolRAG model chooses relevant tools at each step. Training data was synthesized by multi-agent pipelines (ToolGen, QuestionGen, TraceGen). On five new benchmarks covering drug reasoning and personalized treatment, the compact agent beats much larger systems, including GPT-4o, Llama-3.1-70B-Instruct and DeepSeek-R1. Its accuracy stays stable whether a drug is named by brand, generic name or description. The work shows that grounding a small model in verified, updatable tools can matter more than scale for clinical decision support.

## Key points
- **ToolUniverse** contains 211 tools. ToolGen, a multi-agent system, builds such tools automatically from API documentation.
- **TxAgent-Instruct** holds 378,027 instruction-tuning samples. These were derived from 85,340 multi-step traces, which include 177,626 reasoning steps and 281,695 function calls.
- **Benchmarks:** DrugPC, BrandPC, GenericPC, DescriptionPC and TreatmentPC together cover 3,168 drug-reasoning tasks and 456 patient treatment scenarios.
- **Open-ended DrugPC accuracy:** TxAgent scores 92.1%, versus 66.3% for GPT-4o and 52.8% for Llama-3.1-70B-Instruct. On TreatmentPC it leads GPT-4o by 13.6 points, and in open-ended queries it is 7.5 points more accurate than DeepSeek-R1 (671B).
- **Robustness and ablations:** accuracy variance across drug-name variants is \<0.01, compared with 9.96 for GPT-4o, and TxAgent reaches 56.5% on DescriptionPC. Ablations show that three factors improve results: larger toolboxes, real tools rather than LLM-simulated ones, and explicit reasoning before each call. Code is available at github.com/mims-harvard/TxAgent.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

📚 Shelves (5): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 19,503 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 18,805 works · [🧮 Computational Theory and Mathematics](https://md.research.elamur.ai/free/en/shelves/computer-science/computational-theory-and-mathematics/) · 496 works · [🩺 Medicine](https://md.research.elamur.ai/free/en/shelves/medicine/) · 288 works · [🩺 Health Informatics](https://md.research.elamur.ai/free/en/shelves/medicine/health-informatics/) · 227 works

🧭 Topics (1): [🧮 Reasoning](https://md.research.elamur.ai/free/en/topics/reasoning/) · 1,948 works

## Cite

Cite as: Shanghua Gao, Richard Zhu, Zhenglun Kong, Ayush Noori, Xiaorui Su, Curtis Ginder, Theodoros Tsiligkaridis and Marinka Zitnik. 2025. TxAgent: An AI Agent for Therapeutic Reasoning Across a Universe of Tools. arXiv:2503.10970.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@misc{gao2025txagent,
  title = {{TxAgent: An AI Agent for Therapeutic Reasoning Across a Universe of Tools}},
  author = {Shanghua Gao, Richard Zhu, Zhenglun Kong, Ayush Noori, Xiaorui Su, Curtis Ginder, Theodoros Tsiligkaridis and Marinka Zitnik},
  year = {2025},
  doi = {10.48550/arxiv.2503.10970},
  eprint = {2503.10970},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2503.10970}
}
```

- [← previous Towards Holistic Evaluation of Large Audio-Language Models: A Comprehensive Survey](https://md.research.elamur.ai/free/en/papers/yang-2025-holistic-evaluation-large-audio-language/)
- [next → Who Wrote This? The Key to Zero-Shot LLM-Generated Text Detection Is GECScore](https://md.research.elamur.ai/free/en/papers/wu-2025-who-wrote-this-key-zero/)
- [↑ back to papers of 2025](https://md.research.elamur.ai/free/en/papers/2025/page/7/)

## References

24 references, 13 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Huang, K. et al. A foundation model for clinician-centered drug repurposing. Nature Medicine 30, 3601–3613 (2024). \#\# 2. Dubey, A. et al. [The llama 3 herd of models](https://md.research.elamur.ai/free/en/papers/grattafiori-2024-llama-3-herd-models/). arXiv preprint arXiv:2407.21783 (2024).
2. Radford, A., Narasimhan, K., Salimans, T., Sutskever, I. et al. Improving language understanding by generative pre-training (2018). \#\# 4. Jiang, A. Q. et al. Mistral 7b. arXiv preprint arXiv:2310.06825 (2023). [find](https://scholar.google.com/scholar?q=Radford%2C+A%2E%2C+Narasimhan%2C+K%2E%2C+Salimans%2C+T%2E%2C+Sutskever%2C+I%2E+et+al%2E+Improving+language+understanding+by+generative+pre%2Dtraining+%282018%29%2E+%5C%23%5C%23+4%2E+Jiang%2C+A%2E+Q%2E+et+al%2E+Mistral+7b%2E+arXiv+preprint+arXiv%3A2310%2E06)
3. Touvron, H. et al. [Llama: Open and efficient foundation language models](https://md.research.elamur.ai/free/en/papers/touvron-2023-llama-open-efficient-foundation-language/). arXiv preprint arXiv:2302.13971 (2023). \#\# 6. Bai, J. et al. Qwen technical report. arXiv preprint arXiv:2309.16609 (2023).
4. Singhal, K. et al. [Large language models encode clinical knowledge](https://md.research.elamur.ai/free/en/papers/singhal-2022-large-language-models-encode-clinical/). Nature 620, 172–180 (2023).
5. Singhal, K. et al. Toward expert-level medical question answering with large language models. Nature Medicine 1–8 (2025). [find](https://scholar.google.com/scholar?q=Singhal%2C+K%2E+et+al%2E+Toward+expert%2Dlevel+medical+question+answering+with+large+language+models%2E+Nature+Medicine+1%E2%80%938+%282025%29%2E)
6. Chen, Z. et al. [Meditron-70b: Scaling medical pretraining for large language models](https://md.research.elamur.ai/free/en/papers/chen-2023-meditron-70b-scaling-medical-pretraining/). arXiv preprint arXiv:2311.16079 (2023).
7. Alber, D. A. et al. Medical large language models are vulnerable to data-poisoning attacks. Nature Medicine 1–9 (2025). [find](https://scholar.google.com/scholar?q=Alber%2C+D%2E+A%2E+et+al%2E+Medical+large+language+models+are+vulnerable+to+data%2Dpoisoning+attacks%2E+Nature+Medicine+1%E2%80%939+%282025%29%2E)
8. Yan, F. et al. Berkeley function calling leaderboard. [https://gorilla.cs.berkeley.edu/blogs/8](https://gorilla.cs.berkeley.edu/blogs/8) b erkeley function calling leaderboard.html (2024). [↗](https://gorilla.cs.berkeley.edu/blogs/8)
9. Dadao, I. watt-tool-8b: A fine-tuned language model for tool usage and multi-turn dialogue. [https://huggingface.co/watt-ai/watt-tool-8B](https://huggingface.co/watt-ai/watt-tool-8B) (2025). [↗](https://huggingface.co/watt-ai/watt-tool-8B)
10. Liu, W. et al. [Toolace: Winning the points of llm function calling](https://md.research.elamur.ai/free/en/papers/liu-2024-toolace-winning-points-llm-function/). arXiv:2409.00920 (2024). arXiv preprint
11. Gao, Y. et al. [Retrieval-augmented generation for large language models: A survey](https://md.research.elamur.ai/free/en/papers/gao-2023-retrieval-augmented-generation-large-language/). arXiv preprint arXiv:2312.10997 (2023).
12. Gao, S. et al. [Empowering biomedical discovery with ai agents](https://md.research.elamur.ai/free/en/papers/gao-2024-empowering-biomedical-discovery-ai-agents/). Cell 187, 6125–6151 (2024).
13. Boiko, D. A., MacKnight, R., Kline, B. & Gomes, G. [Autonomous chemical research with large language models](https://md.research.elamur.ai/free/en/papers/boiko-2023-autonomous-chemical-research-large-language/). Nature 624, 570–578 (2023).
14. Yao, S. et al. [React: Synergizing reasoning and acting in language models](https://md.research.elamur.ai/free/en/papers/yao-2022-react-synergizing-reasoning-acting-language/). In International Conference on Learning Representations (ICLR) (2023).
15. Bran, A. M. et al. [Chemcrow: Augmenting large-language models with chemistry tools](https://md.research.elamur.ai/free/en/papers/bran-2023-chemcrow-augmenting-large-language/). arXiv preprint arXiv:2304.05376 (2023).
16. Swanson, K., Wu, W., Bulaong, N. L., Pak, J. E. & Zou, J. The virtual lab: Ai agents design new sars-cov-2 nanobodies with experimental validation. bioRxiv 2024–11 (2024). [find](https://scholar.google.com/scholar?q=Swanson%2C+K%2E%2C+Wu%2C+W%2E%2C+Bulaong%2C+N%2E+L%2E%2C+Pak%2C+J%2E+E%2E+%26+Zou%2C+J%2E+The+virtual+lab%3A+Ai+agents+design+new+sars%2Dcov%2D2+nanobodies+with+experimental+validation%2E+bioRxiv+2024%E2%80%9311+%282024%29%2E)
17. Kass-Hout, T. A. et al. Openfda: an innovative platform providing access to a wealth of fda’s publicly available data. Journal of the American Medical Informatics Association 23, 596–600 (2016). [find](https://scholar.google.com/scholar?q=Kass%2DHout%2C+T%2E+A%2E+et+al%2E+Openfda%3A+an+innovative+platform+providing+access+to+a+wealth+of+fda%E2%80%99s+publicly+available+data%2E+Journal+of+the+American+Medical+Informatics+Association+23%2C+596%E2%80%93600+%282016%29%2E)
18. Ochoa, D. et al. The next-generation open targets platform: reimagined, redesigned, rebuilt. Nucleic acids research 51, D1353–D1359 (2023). [find](https://scholar.google.com/scholar?q=Ochoa%2C+D%2E+et+al%2E+The+next%2Dgeneration+open+targets+platform%3A+reimagined%2C+redesigned%2C+rebuilt%2E+Nucleic+acids+research+51%2C+D1353%E2%80%93D1359+%282023%29%2E)
19. Castellanos, F. et al. The human phenotype ontology in 2024: phenotypes around the world. Nucleic Acids Research 52 (2024). \#\# 23. OpenAI. Gpt-4o (2024). [find](https://scholar.google.com/scholar?q=Castellanos%2C+F%2E+et+al%2E+The+human+phenotype+ontology+in+2024%3A+phenotypes+around+the+world%2E+Nucleic+Acids+Research+52+%282024%29%2E+%5C%23%5C%23+23%2E+OpenAI%2E+Gpt%2D4o+%282024%29%2E)
20. Gallifant, J. et al. Language models are surprisingly fragile to drug names in biomedical benchmarks. arXiv preprint arXiv:2406.12066 (2024). [find](https://scholar.google.com/scholar?q=Gallifant%2C+J%2E+et+al%2E+Language+models+are+surprisingly+fragile+to+drug+names+in+biomedical+benchmarks%2E+arXiv+preprint+arXiv%3A2406%2E12066+%282024%29%2E)
21. Chen, C. et al. Clinicalbench: Can llms beat traditional ml models in clinical prediction? arXiv preprint arXiv:2411.06469 (2024). — [arXiv:2411.06469](https://arxiv.org/abs/2411.06469) [find](https://scholar.google.com/scholar?q=Chen%2C+C%2E+et+al%2E+Clinicalbench%3A+Can+llms+beat+traditional+ml+models+in+clinical+prediction%3F+arXiv+preprint+arXiv%3A2411%2E06469+%282024%29%2E+%E2%80%94+%5BarXiv%3A2411%2E06469%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2411%2E06469%29)
22. Guo, D. et al. [Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning](https://md.research.elamur.ai/free/en/papers/deepseek-ai-2025-deepseek-r1-incentivizing-reasoning-capability/). arXiv preprint arXiv:2501.12948 (2025).
23. Lewis, P. et al. [Retrieval-augmented generation for knowledge-intensive nlp tasks](https://md.research.elamur.ai/free/en/papers/lewis-2020-retrieval-augmented-generation-knowledge-intensive/). Advances in Neural Information Processing Systems 33, 9459–9474 (2020).
24. Chandak, P., Huang, K. & Zitnik, M. Building a knowledge graph to enable precision medicine. Nature Scientific Data (2023). \#\# 29. Ji, C. C.-J. et al. Gorilla openfunctions v2 (2024). \#\# 30. OpenAI. Introducing openai o1-preview (2024). 31. Gloss, D., Moxley, R. T. I., Ashwal, S. & Oskoui, M. Practi… — [Attention Is All You Need](https://md.research.elamur.ai/free/en/papers/vaswani-2017-attention-all-need/)

*Updated 3 November 2026*