3rd International Conference on Chemo and BioInformatics, Kragujevac, September 25-26. 2025. (pp. 340-354)
АУТОР(И) / AUTHOR(S): Sofija Stanojlović
Download Full Pdf 
DOI: 10.46793/ICCBIKG25.340S
САЖЕТАК / ABSTRACT:
This study explored different machine learning approaches for predicting the biological activity of molecules against the enzyme 11β-hydroxysteroid dehydrogenase type 2 (11β-HSD2). This enzyme plays a key role in regulating cortisol levels, and its inhibition is associated with the development of certain types of cancer. The aim of the research was to identify molecules with high biological activity that could serve as potential inhibitors of this enzyme. The data used in the study were obtained from the ChEMBL database, while the prediction of biological activity (pIC50) was carried out using regression models based on Lipinski and PADEL descriptors. Through hyperparameter optimization with grid search, the most accurate models for each approach were identified. The best performance was achieved using PADEL descriptors, where the LGBM model obtained an R² of 0.65, MAE of 0.5, and MSE of 0.57. For Lipinski descriptors, the best model was Random Forest, with an R² of 0.51, MAE of 0.6, and MSE of 0.8. Additionally, an encoder-decoder architecture was examined, achieving an R² of 0.5, which indicates its potential for this type of prediction. The results show that PADEL descriptors provide better performance compared to Lipinski descriptors. However, none of the models achieved high precision, suggesting the need to expand the dataset and further optimize the models. This research lays the foundation for applying advanced methods, such as Graph Neural Networks (GNNs), to further investigate key molecular features relevant to biological activity. Such an approach has the potential to accelerate drug discovery and the selection of pharmaceutical candidates, enabling the development of new molecules with optimal properties.
КЉУЧНЕ РЕЧИ / KEYWORDS:
Machine learning, 11β-HSD2, PADEL descriptors, Biological activity (pIC50), Drug discovery
ПРОЈЕКАТ / ACKNOWLEDGEMENT:
ЛИТЕРАТУРА / REFERENCES:
- E. M. Sevick, R.K. Jain. Viscous resistance to blood flow in solid tumors: Effect of hematocrit on intratumor blood viscosity. Cancer Research, 49 (1989) 3513–3519.
- N. Cirillo, D.J. Morgan, M.C. Pedicillo, A. Celentano, L. Lo Muzio, M.J. McCullough, S.S. Prime. Characterisation of the cancer-associated glucocorticoid system: key role of 11b-hydroxysteroid dehydrogenase type 2. 2017. Dostupno na: https://www.nature.com/articles/bjc2017243.pdf
- Z. Krozowski. The type I and type II 11b-hydroxysteroid dehydrogenase enzymes. 1999. Dostupno na: https://pubmed.ncbi.nlm.nih.gov/10419017/#:~:text=The%20type%20I%20isoform%20(11beta,producing%20inactive%2011%2Dketo%20metabolites
- S. Aykul, E. Martinez-Hackert. Determination of half-maximal inhibitory concentration using biosensor-based protein interaction analysis. 2016. Dostupno na: https://pubmed.ncbi.nlm.nih.gov/27365221/#:~:text=Half%2Dmaximal%20inhibitory%20concentration%2 0(IC50,antagonist%20drug%20in%20pharmacological%20research
- M. Nerve. Why using pIC50 instead of IC50 will change your life. 2019. Dostupno na: https://www.collaborativedrug.com/cdd-blog/why-using-pic50-instead-of-ic50-will-change-your-life
- F. Bin Ashraf, S. Akter, S.H. Mumu, M.U. Islam, J. Uddin. Bio-activity prediction of drug candidate compounds targeting SARS-Cov-2 using machine learning approaches. 2023. Dostupno na: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10479925/pdf/pone.0288053.pdf
- Z. Qin, Y. Xi, S. Zhang, G. Tu, A. Yan. Classification of Cyclooxygenase-2 Inhibitors Using Support Vector Machine and Random Forest Methods. 2019. Dostupno na: https://pubs.acs.org/doi/10.1021/acs.jcim.8b00876
- A. Chugh. MAE, MSE, RMSE, Coefficient of Determination, Adjusted R Squared — Which Metric is Better? 2020. Dostupno na: https://medium.com/analytics-vidhya/mae-mse-rmse-coefficient-of- determination-adjusted-r-squared-which-metric-is-better-cd0326a5697e
- P. Le Coz, G. Bouvenot. Éthique et médicament : faut-il s’indigner du prix de certains médicaments? Bulletin de l’Académie Nationale de Médecine 2023;207:642–648. https://doi.org/10.1016/j.banm.2023.03.010
- F. Kaguelidou, M. Ouèdraogo, J-M. Treluyer, C. Le Jeunne, M. Annereau, P. Blanc, et al. Développement des médicaments en pédiatrie : défis existants et recommandations. Therapies 2023;78:95–104 https://doi.org/10.1016/j.therap.2022.12.002
- D. Deplanque, C. Fetro, A. Ferry, P. Lechat, T. Beghyn, C. Bernard, et al. Drug repurposing: From the discovery of a useful pharmacological effect to making the treatment available to the patient. Therapies 2023;78:10–18. https://doi.org/10.1016/j.therap.2022.11.009
- D. Deplanque, C. Fetro, A. Ferry, P. Lechat, T. Beghyn, C. Bernard, et al. Repositionnement des médicaments : de la découverte d’un effet pharmacologique utile à la mise à disposition du traitement pour le patient. Therapies 2023;78:1–9. https://doi.org/10.1016/j.therap.2022.12.001
- X. Zhang, L. Li, M.K. Ng, S. Zhang. Drug–target interaction prediction by integrating multiview network data. Computational Biology and Chemistry 2017;69:185–193. https://doi.org/10.1016/j.compbiolchem.2017.03.011
- O. Harrouche, Y. MBA. Recommender system based on Convolutional Recurrent Deep Learning for protein-drug interaction prediction. Expert Systems with Applications 2024;245:123090. https://doi.org/10.1016/j.eswa.2023.123090
- S. Wang, P. Shan, Y. Zhao, L. Zuo. GanDTI: A multi-task neural network for drug-target interaction prediction. Computational Biology and Chemistry 2021;92:107476. https://doi.org/10.1016/j.compbiolchem.2021.107476
- P. Carracedo-Reboredo, J. Liñares-Blanco, N. Rodríguez-Fernández, F. Cedrón, F.J. Novoa, A. Carballal, et al. A review on machine learning approaches and trends in drug discovery. Computational and Structural Biotechnology Journal 2021;19:4538–4558. https://doi.org/10.1016/j.csbj.2021.08.011
- A. Stroobants, L.H. Mervin, O. Engkvist, G.R. Robb. An industrial evaluation of proteochemometric modelling: Predicting drug-target affinities for kinases. Artificial Intelligence in the Life Sciences 2023;4:100079. https://doi.org/10.1016/j.ailsci.2023.100079
- M. Li, W. Tang, X. Wang, Y. Zhan, B. Chen. Prediction of Pic50 Bioactive Values Based on Knowledge Priors and Attention Neural Networks. 2023. https://doi.org/10.2139/ssrn.4576607
- Z. Liang, H. Li, X. Lu, G. Lin, Y. Li, R. Zhang. 3D-QSAR, in vitro assay and MD simulations studies on the design, bioactivities and different inhibitory modes of the novel DPP-IV inhibitory peptides. Journal of Molecular Structure 2023;1283:135271. https://doi.org/10.1016/j.molstruc.2023.135271
- Artificial Intelligence for Prediction of Biological Activities and Generation of molecular hits using Stereochemical Information. 2023. https://doi.org/10.21203/rs.3.rs-2499317/v1
- S. A. Khedekar, N.S. Mhatre, R. Mendhe. Prediction of pIC50 Values for the Acetylcholinesterase (AChE) using QSAR Model. 2022;09.
- M. Kumari, N.S. Development of Convolutional Neural Network Based Quantitative Structure Activity-Relationship Model and Chemical Space Analysis of Imidazolopiperazines to Identify Lead Compounds Against Malaria. 2023. https://doi.org/10.2139/ssrn.4394372
- Z. Zhu, Z. Rahman, M. Aamir, S.Z.A. Shah, S. Hamid, A. Bilawal, et al. Insight into TLR4 receptor inhibitory activity via QSAR for the treatment of Mycoplasma pneumonia disease. RSC Adv 2023;13:2057–2069. https://doi.org/10.1039/D2RA06178C
