Applications of In Silico Models to Predict Drug-Induced Liver Injury

Deep Learning Prediction of Drug-Induced Liver Toxicity by Manifold Embedding of Quantum Information of Drug Molecules

PURPOSE

Drug-induced liver injury, or DILI, affects numerous patients and also presents significant challenges in drug development. It has been attempted to predict DILI of a chemical by in silico approaches, including data-driven machine learning models. Herein, we report a recent DILI deep-learning effort that utilized our molecular representation concept by manifold embedding electronic attributes on a molecular surface.

METHODS

Local electronic attributes on a molecular surface were mapped to a lower-dimensional embedding of the surface manifold. Such an embedding was featurized in a matrix form and used in a deep-learning model as molecular input. The model was trained by a well-curated dataset and tested through cross-validations.

RESULTS

Our DILI prediction yielded superior results to the literature-reported efforts, suggesting that manifold embedding of electronic quantities on a molecular surface enables machine learning of molecular properties, including DILI.

CONCLUSIONS

The concept encodes the quantum information of a molecule that governs intermolecular interactions, potentially facilitating the deep-learning model development and training.

Utility of human cytochrome P450 inhibition data in the assessment of drug-induced liver injury

Drug-induced liver injury (DILI) is a major cause of drug development discontinuation and drug withdrawal from the market, but there are no golden standard methods for DILI risk evaluation. Since we had found the association between DILI and CYP1A1 or CYP1B1 inhibition, we further evaluated the utility of cytochrome P450 (P450) inhibition assay data for DILI risk evaluation using decision tree analysis.The inhibitory activity of drugs with DILI concern (DILI drugs) and no DILI concern (no-DILI drugs) against 10 human P450s was assessed using recombinant enzymes and luminescent substrates. The drugs were also subjected to cytotoxicity assays and high-content analysis using HepG2 cells. Molecular descriptors were calculated by alvaDesc.Decision tree analysis was performed with the data obtained as variables with or without P450-inhibitory activity to discriminate between DILI drugs and no-DILI drugs. The accuracy was significantly higher when P450-inhibitory activity was included. After the decision tree discrimination, the drugs were further discriminated with the P450-inhibitory activity. The results demonstrated that many false-positive and false-negative drugs were correctly discriminated by using the P450 inhibition data.These results suggest that P450 inhibition assay data are useful for DILI risk evaluation.

In silico modeling-based new alternative methods to predict drug and herb-induced liver injury: A review

New approach methods (NAMs) have been developed to predict a wide range of toxicities through innovative technologies. Liver injury is one of the most extensively studied endpoints due to its severity and frequency, occurring among populations that consume drugs or dietary supplements. In this review, we focus on recent developments of in silico modeling for liver injury prediction using deep learning and in vitro data based on adverse outcome pathways (AOPs). Despite these models being mainly developed using datasets generated from drug-like molecules, they were also applied to the prediction of hepatotoxicity caused by herbal products. As deep learning has achieved great success in many different fields, advanced machine learning algorithms have been actively applied to improve the accuracy of in silico models. Additionally, the development of liver AOPs, combined with big data in toxicology, has been valuable in developing in silico models with enhanced predictive performance and interpretability. Specifically, one approach involves developing structure-based models for predicting molecular initiating events of liver AOPs, while others use in vitro data with structure information as model inputs for making predictions. Even though liver injury remains a difficult endpoint to predict, advancements in machine learning algorithms and the expansion of in vitro databases with relevant biological knowledge have made a huge impact on improving in silico modeling for drug-induced liver injury prediction.

The 3Rs in Experimental Liver Disease

Patients with cirrhosis present multiple physiological and immunological alterations that play a very important role in the development of clinically relevant secondary complications to the disease. Experimentation in animal models is essential to understand the pathogenesis of human diseases and, considering the high prevalence of liver disease worldwide, to understand the pathophysiology of disease progression and the molecular pathways involved, due to the complexity of the liver as an organ and its relationship with the rest of the organism. However, today there is a growing awareness about the sensitivity and suffering of animals, causing opposition to animal research among a minority in society and some scientists, but also about the attention to the welfare of laboratory animals since this has been built into regulations in most nations that conduct animal research. In 1959, Russell and Burch published the book "The Principles of Humane Experimental Technique", proposing that in those experiments where animals were necessary, everything possible should be done to try to replace them with non-sentient alternatives, to reduce to a minimum their number, and to refine experiments that are essential so that they caused the least amount of pain and distress. In this review, a comprehensive summary of the most widely used techniques to replace, reduce, and refine in experimental liver research is offered, to assess the advantages and weaknesses of available experimental liver disease models for researchers who are planning to perform animal studies in the near future.