Skip to content
DraconiticResearch
Search Sources API
189,280 studies31 sourcesUpdated 29 min ago
Peer-reviewedCC BYnpj Computational Materials

Honegumi: An Interface for Accelerating the Adoption of Bayesian Optimization in the Experimental Sciences

  • Sterling G. Baird1,2,3
  • Andrew Falkowski2
  • Taylor D. Sparks1,2
  1. 1Acceleration Consortium, University of Toronto, Toronto, M5S 3H6, Ontario, Canada
  2. 2Department of Materials Science and Engineering, University of Utah, Salt Lake City, 84108, Utah, USA
  3. 3Department of Mechanical Engineering, Brigham Young University, Provo, 84602, Utah, USA
Open at sourcePDFDOICite

Abstract

Bayesian optimization (BO) is increasingly used for experimental design in materials science, chemistry, and biology, yet existing libraries and underlying concepts can be complex for non-machine-learning experts. To address this barrier, we introduce Honegumi, an interactive tool that simplifies creating advanced Bayesian optimization scripts. Honegumi provides a dynamic selection grid for configuring options and automatically generates ready-to-use, unit-tested Python scripts tailored to specific needs. In addition to the interface, a suite of materials-focused tutorials offers both conceptual background and practical guidance, bridging theory and implementation. Built on the Ax platform, Honegumi leverages state-of-the-art BO methods while restructuring the user experience to make advanced techniques more accessible to experimental researchers. By lowering the barrier to entry and providing educational resources, Honegumi aims to accelerate the adoption of advanced BO across scientific domains. The tool is available at https://honegumi.readthedocs.io/ and as a flexible Python package, extensible to other applications.

Related

  1. Accelerating Lead Optimization away from Anti-Targets: Navigating the Structural Complexity of Promiscuous ADMET Proteins
    Howook Hwang, Mee Shelley, Andrew Placzek +22
    ChemRxivPreprintLicence not statedGeneral chemistryChemRxiv
  2. Czochralski Growth of YSGAG Crystals: Experimental, Numerical, and Data-Assisted Analysis of Solid–Liquid Interface Shape
    Kunal Meshram, Dorota Pulmannová, Roberts Blukis +3
    ChemRxivPreprintLicence not statedGeneral chemistryChemRxiv
  3. Bgolearn: a unified Bayesian optimization framework for accelerating materials discovery
    Bin Cao, Jie Xiong, Jiaxuan Ma +11
    npj Computational MaterialsPeer-reviewedCC BY-NC-NDFull text7 citedMaterials & polymersOpenAlex
  4. Roles of oxygen vacancies and the Ni–support interface in CO2 hydrogenation over Ni/Y-doped BaZrO3 revealed by experimental and computational studies
    Tatsuya Joutsuka, Aki Hoshimori, Hiroyuki Yamaura +3
    ChemRxivPreprintCC BY-NC-NDFull textCatalysisChemRxiv
  5. Combining physical models with dynamically acquired experimental information for the optimization of multicomponent NASICON fast ionic conductors in a self-driving laboratory
    Bernardus Rendy, Yuxing Fei, Tanjin He +9
    arXivPreprintCC BY 4.0Full textPhysical chemistryCondensed matterarXiv

Cite

BibTeXRIS
BibTeX
@article{baird2026honegumi,
  title = {{Honegumi: An Interface for Accelerating the Adoption of Bayesian Optimization in the Experimental Sciences}},
  author = {Sterling G. Baird and Andrew Falkowski and Taylor D. Sparks},
  journal = {npj Computational Materials},
  year = {2026},
  month = jun,
  doi = {10.1038/s41524-026-02156-0},
  url = {https://doi.org/10.1038/s41524-026-02156-0},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Honegumi: An Interface for Accelerating the Adoption of Bayesian Optimization in the Experimental Sciences
AU  - Sterling G. Baird
AU  - Andrew Falkowski
AU  - Taylor D. Sparks
PY  - 2026
DA  - 2026/06/03
JO  - npj Computational Materials
DO  - 10.1038/s41524-026-02156-0
UR  - https://doi.org/10.1038/s41524-026-02156-0
AB  - Bayesian optimization (BO) is increasingly used for experimental design in materials science, chemistry, and biology, yet existing libraries and underlying concepts can be complex for non-machine-learning experts. To address this barrier, we introduce Honegumi, an interactive tool that simplifies creating advanced Bayesian optimization scripts. Honegumi provides a dynamic selection grid for configuring options and automatically generates ready-to-use, unit-tested Python scripts tailored to specific needs. In addition to the interface, a suite of materials-focused tutorials offers both conceptual background and practical guidance, bridging theory and implementation. Built on the Ax platform, Honegumi leverages state-of-the-art BO methods while restructuring the user experience to make advanced techniques more accessible to experimental researchers. By lowering the barrier to entry and providing educational resources, Honegumi aims to accelerate the adoption of advanced BO across scientific domains. The tool is available at https://honegumi.readthedocs.io/ and as a flexible Python package, extensible to other applications.
SN  - 2057-3960
ER  -