ORIGINAL RESEARCH article
Front. Chem.
Sec. Theoretical and Computational Chemistry
This article is part of the Research TopicDesign of Extended Networks for Tuning Functionality of MaterialsView all 5 articles
Prediction of cytotoxicity and drug delivery performance of metal organic frameworks through artificial intelligence modeling and validation
Provisionally accepted
- 1University of Hail, Ha'il, Saudi Arabia
- 2University of Hail, Hail, Saudi Arabia
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
This study develops a predictive framework for modeling Drug Loading Capacity as well as Cell Viability (cytotoxicity), two key outputs influenced by molecular and physical properties. Machine learning models were developed and validated to analyze the properties of Metal Organic Frameworks (MOFs) as advanced porous materials for drug delivery. We utilized a Bagging ensemble model built on three regression techniques: Kernel Ridge Regression (KRR), Polynomial Regression (PR), and Poisson Regression (POR). To prepare the data, outlier detection was performed using the Isolation Forest algorithm, identifying and removing outliers for both outputs. Min-Max scaling normalized the data, ensuring all features contributed proportionally to the model. Hyperparameter tuning was performed utilizing the Red Deer Algorithm (RDA), a metaheuristic optimization method aimed at effective parameter selection. The BAG-KRR model constantly outperformed other approaches and produced very remarkable results. This model obtained a Monte Carlo cross-validation (CV) R² of 0.99039 ± 0.00064, a K-fold CV R² of 0.99607 ± 0.00062, and the minimal test error metrics (RMSE = 0.00961, Max Error = 0.02687) for loading capacity. Similarly, for cell viability, the BAG-KRR model achieved near-perfect predictions, recording a Monte Carlo cross-validation R2 of 0.99562 ± 0.00007 and a K-fold cross-validation R2 of 0.99118 ± 0.00009. Additionally, it obtained low test errors, with a RMSE of 0.65925.
Keywords: Drug loading, Kernel ridge regression, MOFs, Poisson regression, Polynomial regression
Received: 18 Dec 2025; Accepted: 11 Feb 2026.
Copyright: © 2026 Huwaimel and Alqarni. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Bader Huwaimel
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.