Reliable Prediction of Caco-2 Permeability by Supervised Recursive Machine Learning Approaches

Abstract

The heterogeneity of the Caco-2 cell line and differences in experimental protocols for permeability assessment using this cell-based method have resulted in the high variability of Caco-2 permeability measurements. These problems have limited the generation of large datasets to develop accurate and applicable regression models. This study presents a QSPR approach developed on the KNIME analytical platform and based on a structurally diverse dataset of over 4900 molecules. Interpretable models were obtained using random forest supervised recursive algorithms for data cleaning and feature selection. The development of a conditional consensus model based on regional and global regression random forest produced models with RMSE values between 0.43–0.51 for all validation sets. The potential applicability of the model as a surrogate for the in vitro Caco-2 assay was demonstrated through blind prediction of 32 drugs recommended by the International Council for the Harmonization of Technical Requirements for Pharmaceuticals (ICH) for validation of in vitro permeability methods. The model was validated for the preliminary estimation of the BCS/BDDCS class. The KNIME workflow developed to automate new drug prediction is freely available. The results suggest that this automated prediction platform is a reliable tool for identifying the most promising compounds with high intestinal permeability during the early stages of drug discovery.

Denoising Drug Discovery Data for Improved ADMET Property Prediction

“…Four ADMET assay datasets were collected from the literature [29][30][31][32][33][34] as listed in Table 1: logD, human Fu,p, Papp, and hERG binding. These assays were chosen because they are important endpoints that are commonly measured in small molecule drug discovery that span a fairly accurate and diverse range of achieved model predictive performances.…”

Harnessing Medicinal Chemical Intuition from Collective Intelligence

“…-In case of duplicates, a single value was assigned per unique compound by keeping the median of the experimental value if the inter-laboratory (SDi) variations did not exceed 0.5 log (Table S1). 71 Machine Learning Models. For the machine learning model, the datasets were divided into training (80%) and test (20%) subsets.…”

Artificial Intelligence and Machine Learning in Pharmacological Research: Bridging the Gap Between Data and Drug Discovery

“…There are several approaches to AI-guided lead optimization which are mentioned as follows: (1) permeability. By exploring ML models, molecular descriptors, and structural patterns, precise predictions for BBB permeability, cytochrome P450 (CYP) enzyme substrate and inhibitor interactions, plasma half-life, solubility, metabolic stability, potential metabolites, renal excretion, bile salt export pump (BSEP) inhibition, hepatotoxicity assessment, and cardiotoxicity can be achieved [18][19][20][21][22][23][24][25][26][27][28][29].…”