Enhancing Response Surface Methodology for Bioprocesses Development Using Interpretable Machine Learning
- 1Universidad Nacional de Colombia sede Manizales , , Campus La Nubia , ,
- 2TBI, Université de Toulouse, CNRS, INRAE, INSA , 135 Avenue de Rangueil , (CEDEX 04) ,
Abstract
Response surface methodology (RSM) has been a cornerstone for experimental bioprocess optimization, yet its reliance on polynomial equations often fails to capture complex nonlinearities and requires extensive experimental data to attempt capturing them. This study evaluates the enhancement of RSM through machine learning (ML) techniques, specifically Gaussian Processes and Support Vector Machines, across three biological systems: succinic acid fermentation, biobutanol production, and continuous ethanol cultivation. To mitigate the “black-box” nature of ML, explainability measures such as Mutual Information (MI) and Local Interpretable Model-agnostic Explanations (LIME) were integrated to assess feature importance and ensure biological relevance. Results demonstrate that ML models consistently outperform traditional polynomials in predictive accuracy, even when having access to less experimental data for calibration and a reduced space of features, particularly on validation data sets where errors were significantly reduced (e.g., reaching MAPE lower than 1% in simulated experimental scenarios). Furthermore, the application of Model-Based Design of Experiments using Latin Hypercube Designs achieved optimal conditions with up to 50% fewer experimental runs compared to traditional factorial approaches. These findings suggest that coupling ML with explainability tools improves the reliability of finding optimal conditions for bioprocess development while also significantly enhancing operational efficiency by reducing experimental runs, time, and costs.
Related
- Optimisation of HS-SPME conditions with a response surface methodology and development of a fast GC/MS method to quantify ethyl acetate in sweet wines
- Mycogenic synthesis of copper oxide nanoparticles using response surface methodology and assessment of their biological potential
- Development of an HPLC method for the determination of phenolic by-products: optimisation of the separation by means of the experimental designs methodology
- Two-Stage Optimization of Oleic Acid Esterification via Response Surface Methodology and Green Chemistry Metrics Selection for Biodiesel from WCO
- Prediction of the properties of sintered copper-graphite composites using response surface methodology
BibTeX
@article{herreraruiz2026enhancing,
title = {{Enhancing Response Surface Methodology for Bioprocesses Development Using Interpretable Machine Learning}},
author = {Juan Federico Herrera-Ruiz and Juan Camilo Castaño-Sánchez and Oscar Andrés Prado-Rubio},
journal = {ACS Omega},
year = {2026},
month = oct,
doi = {10.1021/acsomega.6c09217},
url = {https://pubs.acs.org/acsodf/article/doi/10.1021/acsomega.6c09217/5460713/Enhancing-Response-Surface-Methodology-for},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Enhancing Response Surface Methodology for Bioprocesses Development Using Interpretable Machine Learning
AU - Juan Federico Herrera-Ruiz
AU - Juan Camilo Castaño-Sánchez
AU - Oscar Andrés Prado-Rubio
PY - 2026
DA - 2026/10/29
JO - ACS Omega
DO - 10.1021/acsomega.6c09217
UR - https://pubs.acs.org/acsodf/article/doi/10.1021/acsomega.6c09217/5460713/Enhancing-Response-Surface-Methodology-for
AB - Response surface methodology (RSM) has been a cornerstone for experimental bioprocess optimization, yet its reliance on polynomial equations often fails to capture complex nonlinearities and requires extensive experimental data to attempt capturing them. This study evaluates the enhancement of RSM through machine learning (ML) techniques, specifically Gaussian Processes and Support Vector Machines, across three biological systems: succinic acid fermentation, biobutanol production, and continuous ethanol cultivation. To mitigate the “black-box” nature of ML, explainability measures such as Mutual Information (MI) and Local Interpretable Model-agnostic Explanations (LIME) were integrated to assess feature importance and ensure biological relevance. Results demonstrate that ML models consistently outperform traditional polynomials in predictive accuracy, even when having access to less experimental data for calibration and a reduced space of features, particularly on validation data sets where errors were significantly reduced (e.g., reaching MAPE lower than 1% in simulated experimental scenarios). Furthermore, the application of Model-Based Design of Experiments using Latin Hypercube Designs achieved optimal conditions with up to 50% fewer experimental runs compared to traditional factorial approaches. These findings suggest that coupling ML with explainability tools improves the reliability of finding optimal conditions for bioprocess development while also significantly enhancing operational efficiency by reducing experimental runs, time, and costs.
SN - 2470-1343
ER -