Elamur Research⌕
☰
home / papers / 2024 / ChemReasoner: Heuristic Search over a Large Language Model's Knowledge Space using Quantum-Chemical Feedback

ChemReasoner: Heuristic Search over a Large Language Model's Knowledge Space using Quantum-Chemical Feedback

Paper · 2024 · Henry W. Sprueill, Carl Edwards, Khushbu Agarwal et al.

Summary

Finding better heterogeneous catalysts is central to cleaner chemical processes. The search is hard because it means choosing which chemical descriptors matter and linking them to atomic-scale behaviour, and the number of descriptor combinations grows combinatorially. ChemReasoner treats catalyst discovery as a search problem in an uncertain environment. An LLM (GPT-3.5 or GPT-4) proposes relevant properties and candidate catalysts, expanding a tree of prompts and answers. Each candidate is converted into a 3D catalyst–adsorbate structure and scored by atomistic GNNs trained on DFT data. The scores are based on adsorption energies and, for more complex reactions, reaction pathways and energy barriers. These rewards prune weak branches, so the model's language reasoning is anchored in quantum-chemical evidence rather than taken at face value. A planner chooses search actions without human input and competes well with descriptor lists written by experts. In a CO2-to-methanol case study, the system recovered the commercial Cu/Zn/Al catalyst family. The work demonstrates a reusable recipe for simulation-checked LLM hypothesis generation in science.

Key points

  • The authors are from PNNL, UIUC, and Microsoft Azure Quantum. The paper appeared at ICML 2024 (PMLR 235), and code and data are released at github.com/pnnl/chemreasoner.
  • ChemReasoner-Planner, which uses no human input, outperformed ChemReasoner-Expert, which uses expert-chosen descriptors, in two of the three benchmark query categories.
  • The released resources include expert query benchmarks, provenance trails, more than 700,000 atomistic trajectories, and DFT validation of candidates.
  • In the CO2 hydrogenation case study, the system recommended CuAlZn and also proposed precious-metal bimetallic alloys such as Pd-Au, which favour methanol selectivity.
  • Llama-2 lacked the instruction-following ability the search requires. Another limitation is the reliance on reference 3D structures, which can make DFT verification unreliable.

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

 Fields, topics and concepts

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 19,503 worksComputational Theory and Mathematics · 496 worksPhysics and Astronomy · 70 worksMaterials Science · 58 worksMaterials Chemistry · 58 worksAtomic and Molecular Physics, and Optics · 6 works

🧭 Topics 1Research subjects and languages the work is about.Language models & pretraining · 4,145 works

Cite

Cite as Henry W. Sprueill, Carl Edwards, Khushbu Agarwal, Mariefel V. Olarte, Udishnu Sanyal, Conrad Johnston, Hongbin Liu, Heng Ji and Sutanay Choudhury. 2024. ChemReasoner: Heuristic Search over a Large Language Model's Knowledge Space using Quantum-Chemical Feedback. arXiv:2402.10980.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2402.10980 · arXiv 2402.10980

@misc{sprueill2024chemreasoner,
  title = {{ChemReasoner: Heuristic Search over a Large Language Model's Knowledge Space using Quantum-Chemical Feedback}},
  author = {Henry W. Sprueill, Carl Edwards, Khushbu Agarwal, Mariefel V. Olarte, Udishnu Sanyal, Conrad Johnston, Hongbin Liu, Heng Ji and Sutanay Choudhury},
  year = {2024},
  doi = {10.48550/arxiv.2402.10980},
  eprint = {2402.10980},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2402.10980}
}

References

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

  1. Daza, Y. A. and Kuhn, J. N. Co 2 conversion by reverse water gas shift catalysis: comparison of catalysts, mechanisms and their consequences for co 2 conversion to liquid fuels. RSC advances, 6(55):49675–49691, 2016. find
  2. Artz, J., Müller, T. E., Thenert, K., Kleinekorte, J., Meys, R., Sternberg, A., Bardow, A., and Leitner, W. Sustainable conversion of carbon dioxide: an integrated review of catalysis and life cycle assessment. Chemical reviews, 118(2):434–504, 2018. find
  3. Dumesic, J. A., Huber, G. W., and Boudart, M. Principles of heterogeneous catalysis. Handbook of Heterogeneous Catalysis: Online, 2008. find
  4. Boiko, D. A., MacKnight, R., and Gomes, G. Emergent autonomous scientific research capabilities of large language models. arXiv preprint arXiv:2304.05332, 2023.
  5. Edwards, C., Zhai, C., and Ji, H. Text2Mol: Cross-modal molecule retrieval with natural language queries. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pp. 595–607, Online and Punta Cana, Dominican Republic, November 2021. Association for Computational Li…
  6. Bran, A. M., Cox, S., White, A. D., and Schwaller, P. Chemcrow: Augmenting large-language models with chemistry tools. arXiv preprint arXiv:2304.05376, 2023.
  7. Canakci, M. and Van Gerpen, J. Biodiesel production vi-aacid catalysis. Transactions of the ASAE, 42(5):1203– 1210, 1999. find
  8. Edwards, C., Lai, T., Ros, K., Honke, G., Cho, K., and Ji, H. Translation between molecules and natural language. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pp. 375–413, Abu Dhabi, United Arab Emirates, December 2022a. Association for Computational Lin…
  9. Cao, H., Liu, Z., Lu, X., Yao, Y., and Li, Y. Instructmol: Multi-modal integration for building a versatile and reliable molecular assistant in drug discovery. arXiv preprint arXiv:2311.16208, 2023.
  10. Chanussot, L., Das, A., Goyal, S., Lavril, T., Shuaibi, M., Riviere, M., Tran, K., Heras-Domingo, J., Ho, C., Hu, W., et al. The open catalyst 2020 (oc20) dataset and community challenges. arxiv. arXiv, 2010. find
  11. Edwards, C., Lai, T., Ros, K., Honke, G., Cho, K., and Ji, H. Translation between molecules and natural language. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pp. 375–413, Abu Dhabi, United Arab Emirates, December 2022b. Association for Computational Lin…
  12. Chanussot*, L., Das*, A., Goyal*, S., Lavril*, T., Shuaibi*, M., Riviere, M., Tran, K., Heras-Domingo, J., Ho, C., Hu, W., Palizhati, A., Sriram, A., Wood, B., Yoon, J., Parikh, D., Zitnick, C. L., and Ulissi, Z. Open catalyst 2020 (oc20) dataset and community challenges. find
  13. ACS Catalysis, 2021. doi: 10.1021/acscatal.0c04525. find
  14. Chen, S. and Jung, Y. A generalized-template-based graph neural network for accurate organic reactivity prediction. Nature Machine Intelligence, 4(9):772–780, 2022. find
  15. Etim, U. J., Song, Y., and Zhong, Z. Improving the cu/zno-based catalysts for carbon dioxide hydrogenation to methanol, and the use of methanol as a renewable energy storage media. Frontiers in Energy Research, 8: 545431, 2020. find
  16. Cheng, A. H., Cai, A., Miret, S., Malkomes, G., Phielipp, M., and Aspuru-Guzik, A. Group selfies: a robust fragment-based molecular string representation. Digital Discovery, 2023. find
  17. Fabian, B., Edlich, T., Gaspar, H., Segler, M., Meyers, J., Fiscato, M., and Ahmed, M. Molecular representation learning with language models and domain-relevant auxiliary tasks. arXiv preprint arXiv:2011.13230, 2020. — [arXiv:2011.13230](https://arxiv.org/abs/2011.13230) find
  18. Chithrananda, S., Grand, G., and Ramsundar, B. Chemberta: Large-scale self-supervised pretraining for molecu-Fang, Y., Liang, X., Zhang, N., Liu, K., Huang, R., Chen, Z., Fan, X., and Chen, H. Mol-instructions: A large-scale biomolecular instruction dataset for large language models. arXiv preprint…
  19. Jain, A., Ong, S. P., Hautier, G., Chen, W., Richards, W. D., Dacek, S., Cholia, S., Gunter, D., Skinner, D., Ceder, G., and Persson, K. A. Commentary: The Materials Project: A materials genome approach to accelerating materials innovation. APL Materials, 1(1):011002, 07 2013. ISSN 2166-532X. doi: 1… — [doi:10.1063/1.4812323](https://doi.org/10.1063/1.4812323) find
  20. Gasteiger, J., Becker, F., and Günnemann, S. Gemnet: Universal directional graph neural networks for molecules. Advances in Neural Information Processing Systems, 34: 6790–6802, 2021. find
  21. Kattel, S., Liu, P., and Chen, J. G. Tuning selectivity of co2 hydrogenation reactions at the metal/oxide interface. Journal of the American Chemical Society, 139(29):9739– 9754, 2017. find
  22. Giannozzi, P., Baroni, S., Bonini, N., Calandra, M., Car, R., Cavazzoni, C., Ceresoli, D., Chiarotti, G. L., Cococcioni, M., Dabo, I., et al. Quantum espresso: a modular and open-source software project for quantum simulations of materials. Journal of physics: Condensed matter, 21(39): 395502, 2009. find
  23. Krenn, M., Häse, F., Nigam, A., Friederich, P., and Aspuru-Guzik, A. Self-referencing embedded strings (selfies): A 100% robust molecular string representation. Machine Learning: Science and Technology, 1(4):045024, 2020. find
  24. Giannozzi, P., Andreussi, O., Brumme, T., Bunau, O., Nardelli, M. B., Calandra, M., Car, R., Cavazzoni, C., Ceresoli, D., Cococcioni, M., et al. Advanced capabilities for materials modelling with quantum espresso. Journal of physics: Condensed matter, 29(46):465901, 2017. find
  25. Lai, T. M., Zhai, C., and Ji, H. Knowledge-enhanced biomedical language models. In Journal of Biomedical Informatics, 2023. find
  26. Larsen, A. H., Mortensen, J. J., Blomqvist, J., Castelli, I. E., Christensen, R., Dułak, M., Friis, J., Groves, M. N., Hammer, B., Hargus, C., Hermes, E. D., Jennings, P. C., Jensen, P. B., Kermode, J., Kitchin, J. R., Kolsbjerg, E. L., Kubal, J., Kaasbjerg, K., Lysgaard, S., Marons-son, J. B., Maxs… find
  27. Greeley, J., Nørskov, J. K., and Mavrikakis, M. Electronic structure and catalysis on metal surfaces. Annual review of physical chemistry, 53(1):319–348, 2002. find
  28. Grimme, S., Antony, J., Ehrlich, S., and Krieg, H. A consistent and accurate ab initio parametrization of density functional dispersion correction (dft-d) for the 94 elements h-pu. The Journal of chemical physics, 132(15), 2010. find
  29. Guo, T., Guo, K., Nan, B., Liang, Z., Guo, Z., Chawla, N. V., Wiest, O., and Zhang, X. What can large language models do in chemistry? a comprehensive benchmark on eight tasks. arXiv preprint arXiv:2305.18365, 2023. — [arXiv:2305.18365](https://arxiv.org/abs/2305.18365) find
  30. Ling, G. Z. S., Foo, J. J., Tan, X.-Q., and Ong, W.-J. Transition into net-zero carbon community from fossil fuels: Life cycle assessment of light-driven co2 conversion to methanol using graphitic carbon nitride. ACS Sustainable Chemistry & Engineering, 11(14):5547–5558, 2023. find
  31. Hao, S., Gu, Y., Ma, H., Hong, J. J., Wang, Z., Wang, D. Z., and Hu, Z. Reasoning with language model is planning with world model. arXiv preprint arXiv:2305.14992, 2023. find
  32. Hu, X. and Yip, A. C. Heterogeneous catalysis: enabling a sustainable future, 2021. find
  33. Liu, S., Nie, W., Wang, C., Lu, J., Qiao, Z., Liu, L., Tang, J., Xiao, C., and Anandkumar, A. Multi-modal molecule structure-text model for text-based retrieval and editing. arXiv preprint arXiv:2212.10789, 2022. find
  34. Huang, W., Xia, F., Xiao, T., Chan, H., Liang, J., Florence, P., Zeng, A., Tompson, J., Mordatch, I., Chebotar, Y., et al. Inner monologue: Embodied reasoning through planning with language models. arXiv preprint arXiv:2207.05608, 2022. find
  35. Liu, S., Wang, J., Yang, Y., Wang, C., Liu, L., Guo, H., and Xiao, C. Chatgpt-powered conversational drug editing using retrieval and domain feedback. arXiv preprint arXiv:2305.18090, 2023a. — [arXiv:2305.18090](https://arxiv.org/abs/2305.18090) find
  36. Huber, S. P., Zoupanos, S., Uhrin, M., Talirz, L., Kahle, L., Häuselmann, R., Gresch, D., Müller, T., Yakutovich, A. V., Andersen, C. W., et al. Aiida 1.0, a scalable computational infrastructure for automated reproducible workflows and data provenance. Scientific data, 7(1):300, 2020. find
  37. Liu, S., Zhu, Y., Lu, J., Xu, Z., Nie, W., Gitter, A., Xiao, C., Tang, J., Guo, H., and Anandkumar, A. A text-guided protein design framework. arXiv preprint arXiv:2302.04611, 2023b. — [arXiv:2302.04611](https://arxiv.org/abs/2302.04611) find
  38. Liu, Z., Zhang, W., Xia, Y., Wu, L., Xie, S., Qin, T., Zhang, M., and Liu, T.-Y. Molxpt: Wrapping molecules with text for generative pre-training. arXiv preprint arXiv:2305.10688, 2023c. find
  39. Schwaller, P., Laino, T., Gaudin, T., Bolgar, P., Hunter, C. A., Bekas, C., and Lee, A. A. Molecular transformer: a model for uncertainty-calibrated chemical reaction prediction. ACS central science, 5(9):1572–1583, 2019. find
  40. Mondal, U. and Yadav, G. D. Methanol economy and net zero emissions: critical analysis of catalytic processes, reactors and technologies. Green Chemistry, 23(21):8361– 8405, 2021. find
  41. Schwaller, P., Probst, D., Vaucher, A. C., Nair, V. H., Kreutter, D., Laino, T., and Reymond, J.-L. Mapping the space of chemical reactions using attention-based neural networks. Nature Machine Intelligence, 3(2):144–152, 2021. find
  42. Mukhtar, A., Saqib, S., Lin, H., Shah, M. U. H., Ullah, S., Younas, M., Rezakazemi, M., Ibrahim, M., Mahmood, A., Asif, S., et al. Current status and challenges in the heterogeneous catalysis for biodiesel production. Renewable and Sustainable Energy Reviews, 157:112012, 2022. find
  43. Seidl, P., Vall, A., Hochreiter, S., and Klambauer, G. Enhancing activity prediction models in drug discovery with the ability to understand human language. arXiv preprint arXiv:2303.03363, 2023. — [arXiv:2303.03363](https://arxiv.org/abs/2303.03363) find
  44. Nørskov, J. K., Abild-Pedersen, F., Studt, F., and Bligaard, T. Density functional theory in surface chemistry and catalysis. Proceedings of the National Academy of Sciences, 108(3):937–943, 2011. find
  45. Sprueill, H. W., Edwards, C., Olarte, M. V., Sanyal, U., Ji, H., and Choudhury, S. Monte carlo thought search: Large language model querying for complex scientific reasoning in catalyst design. arXiv preprint arXiv:2310.14420, 2023. find
  46. NVIDIA Corporation. Megamolbart v0.2, 2022. URL https://catalog.ngc.nvidia.com/ orgs/nvidia/teams/clara/models/ megamolbart_0_2. ↗
  47. Stolarczyk, J. K., Bhattacharyya, S., Polavarapu, L., and Feldmann, J. Challenges and prospects in solar water splitting and co2 reduction with inorganic and hybrid nanostructures. ACS Catalysis, 8(4):3602–3635, 2018. find
  48. OpenAI. Gpt-4 technical report, 2023.
  49. Su, B., Du, D., Yang, Z., Zhou, Y., Li, J., Rao, A., Sun, H., Lu, Z., and Wen, J.-R. A molecular multimodal foundation model associating molecule graphs with natural language. arXiv preprint arXiv:2209.05481, 2022. — [arXiv:2209.05481](https://arxiv.org/abs/2209.05481) find
  50. Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., et al. Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35:27730–27744, 2022.
  51. Tawalbeh, M., Javed, R. M. N., Al-Othman, A., Almomani, F., and Ajith, S. Unlocking the potential of CO 2 hydrogenation into valuable products using noble metal catalysts: A comprehensive review. Environmental Technology & Innovation, 31:103217, 2023. ISSN 2352-1864. doi: https://doi.org/10.1016/j.e… — [doi:10.1016/j.eti.2023.103217](https://doi.org/10.1016/j.eti.2023.103217) ↗
  52. Pakhare, D. and Spivey, J. A review of dry (co2) reforming of methane over noble metal catalysts. Chem. Soc. Rev., 43:7813–7837, 2014. doi: 10.1039/C3CS60395D. URL http://dx.doi.org/10.1039/C3CS60395D. — [doi:10.1039/c3cs60395d](https://doi.org/10.1039/c3cs60395d) ↗
  53. Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., Desmaison, A., Kopf, A., Yang, E., DeVito, Z., Raison, M., Tejani, A., Chilamkurthy, S., Steiner, Taylor, R., Kardas, M., Cucurull, G., Scialom, T., Hartshorn, B., Fang, L., Bai,… — Galactica: A Large Language Model for Science
  54. Curran Associates, Inc., 2019. URL — PyTorch: An Imperative Style, High-Performance Deep Learning Library http://papers.neurips.cc/paper/ Tisseraud, C., Comminges, C., Belin, T., Ahouari, H., 9015-pytorch-an-imperative-style-high-performance-deep-learning-library. Soualah, A., Pouilloux, Y., and Le Valant, A. The pdf. cu–zno synergy in methanol synthesis from CO 2 , pa… — [doi:10.1016/j.jcat.2015.04](https://doi.org/10.1016/j.jcat.2015.04) ↗
  55. Touvron, H., Martin, L., Stone, K., Albert, P., Almahairi, A., Babaei, Y., Bashlykov, N., Batra, S., Bhargava, P., Bhosale, S., et al. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288, 2023.
  56. Xu, H. and Wang, S. Protranslator: zero-shot protein function prediction using textual description. In Research in Computational Molecular Biology: 26th Annual International Conference, RECOMB 2022, San Diego, CA, USA, May 22–25, 2022, Proceedings, pp. 279–294. find
  57. Tysinger, E. P., Rai, B. K., and Sinitskiy, A. V. Can we quickly learn to “translate” bioactive molecules with transformer models? Journal of Chemical Information and Modeling, 63(6):1734–1744, 2023. find
  58. Xu, M., Yuan, X., Miret, S., and Tang, J. Protst: Multimodality learning of protein sequences and biomedical texts. arXiv preprint arXiv:2301.12040, 2023. — [arXiv:2301.12040](https://arxiv.org/abs/2301.12040) find
  59. Xu, S. and Carter, E. A. Theoretical insights into heterogeneous (photo) electrochemical co2 reduction. Chemical reviews, 119(11):6631–6669, 2018. find
  60. Unsleber, J. P., Liu, H., Talirz, L., Weymuth, T., Mörchen, M., Grofe, A., Wecker, D., Stein, C. J., Panyala, A., Peng, B., et al. High-throughput ab initio reaction mechanism exploration in the cloud with automated multi-reference validation. The Journal of Chemical Physics, 158(8), 2023. find
  61. Yao, S., Yu, D., Zhao, J., Shafran, I., Griffiths, T. L., Cao, Y., and Narasimhan, K. Tree of thoughts: Deliberate problem solving with large language models. arXiv preprint arXiv:2305.10601, 2023. find
  62. Vall, A., Hochreiter, S., and Klambauer, G. Bioassayclr: Prediction of biological activity for novel bioassays based on rich textual descriptions. In ELLIS ML4Molecules workshop, 2021. find
  63. Ye, G., Cai, X., Lai, H., Wang, X., Huang, J., Wang, L., Liu, W., and Zeng, X. Drugassist: A large language model for molecule optimization. arXiv preprint arXiv:2401.10334, 2023. — [arXiv:2401.10334](https://arxiv.org/abs/2401.10334) find
  64. Vaucher, A. C., Schwaller, P., Geluykens, J., Nair, V. H., Iuliano, A., and Laino, T. Inferring experimental procedures from text-based representations of chemical reactions. Nature communications, 12(1):2573, 2021. find
  65. Zeng, Z., Yao, Y., Liu, Z., and Sun, M. A deep-learning system bridging molecule structure and biomedical text with comprehension comparable to human professionals. Nature communications, 13(1):862, 2022. find
  66. Wang, Y., Cao, L., Libretto, N. J., Li, X., Li, C., Wan, Y., He, C., Lee, J., Gregg, J., Zong, H., Su, D., Miller, J. T., Mueller, T., and Wang, C. Ensemble effect in bimetallic electrocatalysts for co2 reduction. Journal of the American Chemical Society, 141(42):16635–16642, 2019. doi: 10.1021/jacs… — [doi:10.1021/jacs.9b05766](https://doi.org/10.1021/jacs.9b05766) find
  67. Wei, J., Wang, X., Schuurmans, D., Bosma, M., Chi, E., Le, Q., and Zhou, D. Chain of thought prompting elicits reasoning in large language models. arXiv preprint arXiv:2201.11903, 2022.
  68. Zhang, X., Wang, L., Helwig, J., Luo, Y., Fu, C., Xie, Y., Liu, M., Lin, Y., Xu, Z., Yan, K., et al. Artificial intelligence for science in quantum, atomistic, and continuum systems. arXiv preprint arXiv:2307.08423, 2023. find
  69. Zhao, H., Liu, S., Ma, C., Xu, H., Fu, J., Deng, Z.-H., Kong, L., and Liu, Q. Gimlet: A unified graph-text model for instruction-based molecule zero-shot learning. bioRxiv, pp. 2023–05, 2023a. find
  70. Zhao, W., Zhou, D., Cao, B., Zhang, K., and Chen, J. Adversarial modality alignment network for cross-modal molecule retrieval. IEEE Transactions on Artificial Intelligence, 2023b. find
  71. Weininger, D. Smiles, a chemical language and information system. 1. introduction to methodology and encoding rules. Journal of chemical information and computer sciences, 28(1):31–36, 1988. find
  72. Zhao, Z., Ma, D., Chen, L., Sun, L., Li, Z., Xu, H., Zhu, Z., Zhu, S., Fan, S., Shen, G., et al. Chemdfm: Dialogue foundation model for chemistry. arXiv preprint arXiv:2401.14818, 2024. — [arXiv:2401.14818](https://arxiv.org/abs/2401.14818) find
  73. Weininger, D., Weininger, A., and Weininger, J. L. Smiles. 2. algorithm for generation of unique smiles notation. Journal of chemical information and computer sciences, 29(2):97–101, 1989. find
  74. Zitnick, C. L., Chanussot, L., Das, A., Goyal, S., Heras-Domingo, J., Ho, C., Hu, W., Lavril, T., Palizhati, A., Riviere, M., et al. An introduction to electrocatalyst design using machine learning for renewable energy storage. arXiv preprint arXiv:2010.09435, 2020. find
  75. Xi, J., Wang, Z., Wang, W., and Lu, G. In-situ xps study for reaction mechanism of methanol decomposition over cu-ni/zn catalyst. ACTA PHYSICOCHIMICA SINICA, 18 (1):82–86, 2002. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/sprueill-2024-chemreasoner-heuristic-search-over-large/