Enhancing Molecularly Imprinted Polymer Development with Machine Learning Techniques

Abstract

Molecular imprinted polymers (MIPs) have various applications such as chemical sensing and drug delivery systems because they can selectively attach molecules that fit the template. The imprinting factor (IF) is used to assess the degree of contact between the binding sites on MIPs and target molecules, enabling quantification of the binding affinity for MIPs. The machine learning (ML) prediction was carried out in association with a study conducted with 5-fluorouracil, which is a chemotherapy drug used to fight cancer. ML techniques, including K-Nearest Neighbors, Random Forest, XGBoost, Decision Tree, Lasso, and Ridge Regressions, were used to predict IF values for MIP. The dataset includes 214 observations consisting of various template molecules, functional monomers, and crosslinkers. The performance of the model was evaluated and compared using root mean squared error (RMSE) values, mean absolute error (MAE), and R². The Random Forest shows good performance with moderate RMSE (3.54), low MAE (2.19), and R² = 0.55, indicating acceptable error and moderate variance. A Streamlit-based web interface was developed for visualization of polymer interactions and usability of the models by researchers with similar interests. These results imply that with improved and more extensive data, we have the potential to achieve higher IF prediction accuracy with ML models. The limitations of conventional MIPs make experimental estimation of binding affinity challenging due to cost and time constraints; therefore, this study employs ML as a pre-screening tool for prediction.



Received: 6 October 2025 | Revised: 20 April 2026 | Accepted: 19 June 2026



Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.



Data Availability Statement

Data are available from the corresponding author upon reasonable request.



Author Contribution Statement

Jisha R. C.: Methodology, Software, Formal analysis, Writing – original draft, Writing – review & editing, Visualization. Niranjan Prasanth: Software, Formal analysis, Writing – original draft. Aswathi Perumkunnappala: Investigation, Data curation. Meenakshi Krishnan Praseetha Radhakrishnan: Investigation, Data curation. Manjusha Nair: Conceptualization, Validation, Formal analysis, Resources, Writing – review & editing, Supervision, Project administration. Renuka Suravajhala: Conceptualization, Validation, Investigation, Resources, Writing – review & editing, Supervision, Funding acquisition.