Artificial Intelligence-Based Quantitative Structure–Property Relationship Model for Predicting Human Intestinal Absorption of Compounds with Serotonergic Activity

Abstract

Oral medicines represent the largest pharmaceutical market area. To achieve a therapeutic effect, a drug must penetrate the intestinal walls, the main absorption site for orally delivered active pharmaceutical ingredients (APIs). Indeed, predicting drug absorption can facilitate candidate screening and reduce time to market. Algorithms are available with good prediction accuracy that however focus only on solubility. In this work, we focused on drug permeability looking at human intestinal absorption as a marker for intestinal bioavailability. Being of considerable therapeutic relevance, APIs with serotonergic activity were selected as a dataset. Due to process complexity, experimental data scarcity, and variability, we turned toward an artificial intelligence (AI)-based system, which is a hierarchical combination of classification and regression models. This combination of seemingly two models into a single system widens the space of molecules classified as highly permeable with high accuracy. The specialized and optimized system enables in silico and structure-based prediction with a high degree of certainty. Predictions in external validation allowed correct selection of the 38% of highly permeable molecules without any false positives. The proposed system based on AI represents a promising tool useful for oral drug screening at an early stage of drug discovery and development. Datasets and the obtained models are available on the GitHub platform ().

SerotoninAI: Serotonergic System Focused, Artificial Intelligence-Based Application for Drug Discovery

“…Mordred descriptors were used as a molecular representation of chemical structures . Mordred effectiveness in this field and suitability for interpretation has been tested in our previous work . QSAR and QSPR models were created based on two-dimensional descriptors to prevent the risk of variability in three-dimensional descriptors due to nonconvergent molecules’ optimization.…”

“…This analysis was conducted in Python environment, leveraging a framework developed by Szlęk, further augmented with a wrapper for the mljar package. Summary plots of SHAP analysis for each obtained model are presented in Supporting Information S3, except for the HIA model, where the entire analysis was described in previous article . The values of these descriptors in the training set were transformed via Min–Max normalization.…”

“…For the molecule, a confirmed BBB penetration is a crucial property for drugs acting on the central nervous system, whereas extensive HIA ensures feasibility of the oral administration route, the most preferred and acceptable one in the therapy. The artificial intelligence (AI)-based system for HIA has been developed on data sets with serotonergic ligands, which limits its application to these molecules only …”

Combined data-driven and mechanism-based approaches for human-intestinal-absorption prediction in the early drug-discovery stage

“…From another viewpoint, the published machine learning models completely ignored the dose amount; namely, there were no dose-related items as explanatory variables. [17][18][19][20][21][22][23]25,26 However, for solubility-limited drugs, human Fa may vary according to dosage, 27 necessitating the investigation of the effect of the dosage. A mechanism-based Fa prediction method that converts drug solubility, diffusion, and membrane permeability into a mathematical model, known as the gastrointestinal unied theoretical framework (GUTFW) model, has been published.…”

“…25 In the second study, conducted in 2023, the authors focused on specific compounds with serotonergic activity and built a 2-class classifier (AUC 0.72 using the test set) and a regression model ( R 2 = 0.047 using the test set). 26 Hence, to advance the prediction of human Fa, we aimed to build a quantitative regression structure–activity relationship (QSAR) model without dataset selection. The merit of a regression model is that it can show the exact number as a percentage of Fa, which can be easily understood by users.…”