Democratising real-world drug discovery through agentic AI
Open accessAbstract
• Agentic systems are an emerging technology for automation, but evidence of their real-world use in drug discovery is limited. • We present AstraZeneca’s agentic systems LangDMTA and ChatInvent, outlining implementation challenges and lessons learned. ChatInvent, a conversational interface for agentic drug discovery, is illustrated with a typical use case. • Key technical considerations are discussed, including architecture choice, automated testing for continuous integration, and benchmarking. • Practical recommendations are provided for building agentic systems in drug discovery. • We speculate on the potential of agentic systems for improving efficiency in pharmaceutical R&D. Agentic systems that are based on large language models (LLMs) have emerged as promising tools in the chemistry domain over the past few years. Early examples included work on CoScientist, Chemcrow, and LLM-RDF, which showcased the potential of agentic systems to assist in chemical research, in the orchestration of cheminformatics tools, and in synthetic reaction development. Despite this, the current literature lacks examples of the real-world adoption of such systems in drug discovery. We present such an example by describing our work on an agentic system called ChatInvent, which has been integrated into the discovery pipeline at AstraZeneca to aid in molecular design and synthesis planning. We discuss how the system evolved from a proof-of-concept single agent into an extensible, robust, and scalable multi-agent architecture with a graphical user interface. We emphasize the lessons learnt and the challenges that persist as we continue to work on this project, and share our perspectives on the future of agentic systems in our domain.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Emerging "AI organic chemist" platforms combine synthesis planning algorithms with laboratory automation interfaces, enabling partial or closed-loop experimentation.
“Emerging “AI organic chemist” platforms combine synthesis planning algorithms with laboratory automation interfaces, enabling partial or closed-loop experimentation [34, 47].”
From The next paradigm in bioinformatics: a review of multi-agent systems and foundational models for end-to-end scientific discovery (2026) · cites this paper as [47]
Reported implementations suggest a substantial increase in productivity in screening and triage cycles when reasoning and execution levels are closely correlated.
“Reported implementations suggest a substantial increase in productivity in screening and triage cycles when reasoning and execution levels are closely correlated [45, 47].”
From The next paradigm in bioinformatics: a review of multi-agent systems and foundational models for end-to-end scientific discovery (2026) · cites this paper as [47]
In preclinical development, agentic systems increasingly integrate reasoning cycles with controlled instrument execution.
“In preclinical development, agentic systems increasingly integrate reasoning cycles with controlled instrument execution.”
From The next paradigm in bioinformatics: a review of multi-agent systems and foundational models for end-to-end scientific discovery (2026) · cites this paper as [47]
The value of agentic AI in drug discovery and personalized medicine emerges when reasoning, action, and evidence monitoring operate as an integrated system capable of accelerating insight without compromising reproducibility or governance.
“As recent implementations indicate, the value of agentic AI in drug discovery and personalized medicine emerges when reasoning, action, and evidence monitoring operate as an integrated system capable of accelerating insight without compromising reproducibility or governance [27, 33, 34, 45–48].”
From The next paradigm in bioinformatics: a review of multi-agent systems and foundational models for end-to-end scientific discovery (2026) · cites this paper as [47]
Cites
Autonomous chemical research with large language models
Boiko, D. A., MacKnight, R., Kline, B., & Gomes, G. (2023) · Nature · doi:10.1038/s41586-023-06792-0 · Cited by 1,050 · see citing works
Augmenting large language models with chemistry tools
M. Bran, A., Cox, S., Schilter, O., Baldassari, C., White, A. D., & Schwaller, P. (2024) · Nature Machine Intelligence · doi:10.1038/s42256-024-00832-8 · Cited by 794 · see citing works
ChatGPT Chemistry Assistant for Text Mining and the Prediction of MOF Synthesis
Zheng, Z.-L., Zhang, O.-F., Borgs, C., Chayes, J., & Yaghi, O. (2023) · Journal of the American Chemical Society · doi:10.1021/jacs.3c05819 · Cited by 484 · see citing works
Reinvent 4: Modern AI–driven generative molecule design
Loeffler, H. H., He, J., Tibo, A., Janet, J. P., Voronov, A., Mervin, L., & Engkvist, O. (2024) · Journal of Cheminformatics · doi:10.1186/s13321-024-00812-5 · Cited by 385 · see citing works
Self-citationA review of large language models and autonomous agents in chemistry
Ramos, M. C., Collison, C. J., & White, A. D. (2025) · Chemical Science · doi:10.1039/d4sc03921a · Cited by 274 · see citing works
Artificial Intelligence for Drug Discovery: Are We There Yet?
Hasselgren, C., & Oprea, T. I. (2024) · Annual Review of Pharmacology and Toxicology · doi:10.1146/annurev-pharmtox-040323-040828 · Cited by 236 · see citing works
AI-Driven Drug Discovery: A Comprehensive Review
Ferreira, F. J. N., & Carneiro, A. S. (2025) · ACS Omega · doi:10.1021/acsomega.5c00549 · Cited by 202 · see citing works
Mirza, A., Alampara, N., Kunchapu, S., Ríos-García, M., Emoekabu, B., Krishnan, A., Gupta, T., Schilling-Wilhelmi, M., Okereke, M., Aneesh, A., et al. (2025) · Nature Chemistry · doi:10.1038/s41557-025-01815-x · Cited by 156 · see citing works
An automatic end-to-end chemical synthesis development platform powered by large language models
Ruan, Y., Lu, C., Xu, N., He, Y., Chen, Y.-X., Zhang, J., Xuan, J., Pan, J., Fang, Q., Gao, H., et al. (2024) · Nature Communications · doi:10.1038/s41467-024-54457-x · Cited by 140 · see citing works
Plowright, A. T., Johnstone, C., Kihlberg, J. O., Pettersson, J. A., Robb, G. R., & Thompson, R. A. (2011) · Drug Discovery Today · doi:10.1016/j.drudis.2011.09.012 · Cited by 132 · see citing works
Large language models for chemistry robotics
Yoshikawa, N., Skreta, M., Darvish, K., Arellano-Rubach, S., Ji, Z., Bjørn Kristensen, L., Li, A. Z., Zhao, Y., Xu, H., Kuramshin, A., et al. (2023) · Autonomous Robots · doi:10.1007/s10514-023-10136-2 · Cited by 112 · see citing works
Emergent autonomous scientific research capabilities of large language models
Boiko, D. A., MacKnight, R., & Gomes, G. (2023) · arXiv · doi:10.48550/arxiv.2304.05332 · Cited by 72 · see citing works
AiZynthFinder 4.0: developments based on learnings from 3 years of industrial application
Saigiridharan, L., Hassen, A. K., Lai, H., Torren-Peraire, P., Engkvist, O., & Genheden, S. (2024) · Journal of Cheminformatics · doi:10.1186/s13321-024-00860-x · Cited by 69 · see citing works
Self-citationAugmenting DMTA using predictive AI modelling at AstraZeneca
Ghiandoni, G. M., Evertsson, E., Riley, D. J., Tyrchan, C., & Rathi, P. C. (2024) · Drug Discovery Today · doi:10.1016/j.drudis.2024.103945 · Cited by 44 · see citing works
Self-citationCACTUS: Chemistry Agent Connecting Tool Usage to Science
McNaughton, A. D., Ramalaxmi, G. K. S., Kruel, A., Knutson, C. R., Varikoti, R. A., & Kumar, N. (2024) · ACS Omega · doi:10.1021/acsomega.4c08408 · Cited by 37 · see citing works
SynAsk: unleashing the power of large language models in organic synthesis
Zhang, C., Lin, Q., Zhu, B., Yang, H., Lian, X., Deng, H., Zheng, J., & Liao, K. (2024) · Chemical Science · doi:10.1039/d4sc04757e · Cited by 29 · see citing works
Large language models open new way of AI-assisted molecule design for chemists
Ishida, S., Sato, T., Honma, T., & Terayama, K. (2025) · Journal of Cheminformatics · doi:10.1186/s13321-025-00984-8 · Cited by 26 · see citing works
Vangala, S. R., Krishnan, S. R., Bung, N., Nandagopal, D., Ramasamy, G., Kumar, S., Sankaran, S., Srinivasan, R., & Roy, A. (2024) · Journal of Cheminformatics · doi:10.1186/s13321-024-00928-8 · Cited by 25 · see citing works
Precedent Finder: Locating Pareto-Optimal Reactions
Bauer, C. A., Kogej, T., Genheden, S., & Norrby, P. (2025) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.5c01797 · Cited by 5 · see citing works
Self-citation
Flagged first, then by citations
Cited by
Sorted by citations
Gangwal, A., Ansari, A., Siddique, M. U. M., Sawale, J. A., Padmane, S. P., & Wadekar, R. R. (2026) · Chemical Biology & Drug Design · doi:10.1111/cbdd.70372 · Cited by 6 · see citing works
Branda, F., Ahmed, M. M., Ciccozzi, M., Guzzi, P. H., & Scarpa, F. (2026) · Briefings in Bioinformatics · doi:10.1093/bib/bbag245 · Cited by 2 · see citing works
Organic Chemistry as a Catalyst for AI Innovation: Challenges, Methods, and Emerging Paradigms
Chawla, N. V., González‐Montiel, G. A., Guo, K., Guo, T., Hua, T., Huang, X., Inae, E., Jiang, M., Le, K., Liu, G., et al. (2026) · Chemical Reviews · doi:10.1021/acs.chemrev.5c01081 · Cited by 2 · see citing works
Closed-loop agentic AI in drug discovery
Chhetri, P. (2026) · RSC Medicinal Chemistry · doi:10.1039/d6md00319b · Cited by 2 · see citing works
Catalyst Design for Sustainable Reactions: A Primer toward Knowledge-Driven Innovation
Mendes, P. S. F., Costa, D. P., Lauwaert, J., & Thybaut, J. W. (2026) · Chemical Reviews · doi:10.1021/acs.chemrev.5c01041 · Cited by 1 · see citing works
Artificial intelligence in drug development: From molecular discovery to precision medicine
Xie, T., Zhang, Z., Wang, Y., Wang, H., Chen, X., Feng, H., & Shao, F. (2026) · Genes & Diseases · doi:10.1016/j.gendis.2026.102402 · Cited by 1 · see citing works
ChemLint: Conversational Cheminformatics with Large Language Models
van Tilborg, D., & Grisoni, F. (2026) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.6c00638 · Cited by 1 · see citing works
Zhang, L., Wang, L., Sun, X., Tang, W., Su, H., Qian, Y., Yang, Q., Li, Q., Tang, Z., Sun, H., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.04.03.716272
Zhang, Z.-Y., Savvas, G., Li, J., Zhou, Y., Le, T. T., Brandell, D., Mercado, R., & Zhang, C. (2026) · Macromolecular Materials and Engineering · doi:10.1002/mame.70340
R, M. D., & Pallavi, G. B. (2026) · International Journal of Drug Delivery Technology · doi:10.25258/ijddt.16.54s.157
Hassan, L., Khalaf, O., & Jaber, R. (2026) · International Journal of Pharmaceutical and Phytopharmacological Research · doi:10.51847/5tffxnr14x
Artificial intelligence agents and agentic artificial intelligence applied to precision medicine.
Costa, F. F. (2026) · Drug Discovery Today · doi:10.1016/j.drudis.2026.104781
Entering the agentic era of AI in drug discovery
Vijayan, R. S. K., Cross, J. B., & Poongavanam, V. (2026) · Nature Chemical Biology · doi:10.1038/s41589-026-02295-x
Designing an LLM-based evaluation system for agentic AI in drug discovery through human alignment
Granqvist, E., Mercado, R., & Genheden, S. (2026) · Artificial Intelligence in the Life Sciences · doi:10.1016/j.ailsci.2026.100187
Self-citationMikołajewska, E., Rogalla-Ładniak, U., Masiak, J., Panas, E., & Mikołajewski, D. (2026) · Applied Sciences · doi:10.3390/app16189001
MolecureAI: an agentic AI platform for autonomous drug discovery
Chadi, M.-A., Asmaa, A., Hajar, E., & Mousannif, H. (2026) · Frontiers in Pharmacology · doi:10.3389/fphar.2026.1917774
Can autonomous Agentic-AI systems accelerate drug discovery and development in the next decade?
Chaturvedi, N., Haque, S., & Kumar, V. (2026) · Briefings in Bioinformatics · doi:10.1093/bib/bbag583
Huang, Z., Jiang, Z., Hu, S., Ye, N., Liu, X., Zhan, X., Wu, Z., Chen, H., Zhao, Z., Li, X., et al. (2026) · Frontiers in Pharmacology · doi:10.3389/fphar.2026.1970742
- Authors
- He, J., Lai, H., Saigiridharan, L., Ghiandoni, G. M., Jenei, K., Gokalp, U., Nuković, A., Engkvist, O., Janet, J. P., & Genheden, S.
- Affiliations
- AstraZeneca (Sweden) (SE)
- AstraZeneca (United Kingdom) (GB)
- Corresponding author
- Ghiandoni, Gian Marco
- ORCID
- Venue
- Drug Discovery Today
- Publisher
- Elsevier BV
- Published
- 2026-03
- Volume
- 31
- Issue
- 2
- Pages
- 104605
- ISSN
- 1359-6446 (ISSN-L)
- OpenAlex
- W7124507199
- Cited by
- 17
- References
- 18
- Citations by year
- 2026: 17
- FWCI
- 61.72
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Drug Discovery Today (publishedVersion, cc-by-nc-nd)
- PubMed (publishedVersion)
- Chalmers Research (Chalmers University of Technology) (submittedVersion)
- Funders
- AstraZeneca
- Topics
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Computational Drug Discovery Methods
- Physical Sciences › Materials Science › Materials Chemistry › Machine Learning in Materials Science
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence Applications
- Keywords
- agentic AI, large language models, drug discovery, multi-agent architecture, molecular design, synthesis planning, conversational interfaces, continuous integration testing, benchmarking
- MeSH terms
- Cheminformatics, Large Language Models, Artificial Intelligence, Humans, Drug Design, Drug Discovery
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
