Abstract:Although computational predictions of pharmacokinetics (PK) are desirable at the drug design stage, existing approaches are often limited by prediction accuracy and human interpretability. Using a discovery data set of mouse and rat PK studies at Roche (9,685 unique compounds), we performed a proof-of-concept study to predict key PK properties from chemical structure alone, including plasma clearance (CLp), volume of distribution at steady-state (Vss), and oral bioavailability (F). Ten machine learning (ML) models were evaluated, including Single-Task, Multitask, and transfer learning approaches (i.e., pretraining with in vitro data). In addition to prediction accuracy, we emphasized human interpretability of outcomes, especially the quantification of uncertainty, applicability domains, and explanations of predictions in terms of molecular features. Results show that intravenous (IV) PK properties (CLp and Vss) can be predicted with good precision (average absolute fold error, AAFE of 1.96-2.84 depending on data split) and low bias (average fold error, AFE of 0.98-1.36), with AutoGluon, Gaussian Process Regressor (GP), and ChemProp displaying the best performance. Driven by higher complexity of oral PK studies, predictions of F were more challenging, with the best AAFE values of 2.35-2.60 and higher overprediction bias (AFE of 1.45-1.62). Multi-Task approaches and pretraining of ChemProp neural networks with in vitro data showed similar precision to Single-Task models but helped reduce the bias and increase correlations between observations and predictions. A combination of GP-computed prediction variance, molecular clustering, and dimensionality-reduction provided valuable quantitative insights into prediction uncertainty and applicability domains. SHAPley Additive exPlanations (SHAPs) highlighted molecular features contributing to prediction outcomes of Vss, providing explanations that could aid drug design. Combined results show that computational predictions of PK are feasible at the drug design stage, with several ML technologies converging to successfully leverage historical PK data sets. Further studies are needed to unlock the full potential of this approach, especially with respect to data set sizes and quality, transfer learning between in vitro and in vivo data sets, model-independent quantification of uncertainty, and explainability of predictions.

In silico prediction of pharmaceutical degradation pathways: a benchmarking study.

In Silico Prediction of Pharmaceutical Degradation Pathways: A Benchmarking Study Using the Software Program Zeneth

Drug Degradation Prediction, In Silico Toxicity Assessment and Development of Stability-Indicating, Quality by Design Enabled UFLC Method for Sacubitril-Valsartan

Early clinical drug product shelf-life setting using accelerated predictive stability and metabolite data for impurity qualification: A case study

Insight into in silico prediction and chemical degradation study of osimertinib mesylate by LC-HRMS and NMR: Investigation of a typical case of alkaline pH-mediated oxidative degradation product

In Silico Prediction of N -Nitrosamine Formation Pathways of Pharmaceutical Products

In silico prediction of metabolic stability for ester-containing molecules: Machine learning and quantum mechanical methods

Stability Modeling Methodologies to Enable Earlier Patient Access

Computational Predictions of Nonclinical Pharmacokinetics at the Drug Design Stage

Unveiling the potential of machine learning in cost-effective degradation of pharmaceutically active compounds: A stirred photo-reactor study

Entering the era of computationally driven drug development

A Perspective on the Prediction of Drug Pharmacokinetics and Disposition in Drug Research and Development

Predicting deamidation and isomerization sites in therapeutic antibodies using structure-based in silico approaches

Actionable predictions of human pharmacokinetics at the drug design stage

Evaluation of the Success of High-Throughput Physiologically Based Pharmacokinetic (HT-PBPK) Modeling Predictions to Inform Early Drug Discovery

Degradation and Impurity Analysis for Pharmaceutical Drug Candidates

RD-Metabolizer: an Integrated and Reaction Types Extensive Approach to Predict Metabolic Sites and Metabolites of Drug-Like Molecules

In vitro biological studies and computational prediction-based analyses of pyrazolo[1,5- a ]pyrimidine derivatives

AiZynthFinder 4.0: developments based on learnings from 3 years of industrial application

Predicting the price of molecules using their predicted synthetic pathways

Shared learning from a physiologically based pharmacokinetic modeling strategy for human pharmacokinetics prediction through retrospective analysis of Genentech compounds