Predicting Drug-Induced Liver Injury Using Convolutional Neural Network and Molecular Fingerprint-Embedded Features

© 2020 American Chemical Society. As a critical issue in drug development and postmarketing safety surveillance, drug-induced liver injury (DILI) leads to failures in clinical trials as well as retractions of on-market approved drugs. Therefore, it is important to identify DILI compounds in the early-stages through in silico and in vivo studies. It is difficult using conventional safety testing methods, since the predictive power of most of the existing frameworks is insufficiently effective to address this pharmacological issue. In our study, we employ a natural language processing (NLP) inspired computational framework using convolutional neural networks and molecular fingerprint-embedded features. Our development set and independent test set have 1597 and 322 compounds, respectively. These samples were collected from previous studies and matched with established chemical databases for structural validity. Our study comes up with an average accuracy of 0.89, Matthews's correlation coefficient (MCC) of 0.80, and an AUC of 0.96. Our results show a significant improvement in the AUC values compared to the recent best model with a boost of 6.67%, from 0.90 to 0.96. Also, based on our findings, molecular fingerprint-embedded featurizer is an effective molecular representation for future biological and biochemical studies besides the application of classic molecular fingerprints.

Download Full-text

Drug-induced liver injury classification model based on in vitro human transcriptomics and in vivo rat clinical chemistry data

Systems Biomedicine ◽

10.4161/sysb.29400 ◽

2014 ◽

Vol 2 (4) ◽

pp. 63-70 ◽

Cited By ~ 10

Author(s):

Danyel Jennen ◽

Jan Polman ◽

Mark Bessem ◽

Maarten Coonen ◽

Joost van Delft ◽

...

Keyword(s):

Liver Injury ◽

Clinical Chemistry ◽

Classification Model ◽

Drug Induced ◽

Drug Induced Liver Injury ◽

Chemistry Data ◽

Model Based ◽

Injury Classification

Download Full-text

Lighting up alkaline phosphatase in drug-induced liver injury using a new chemiluminescence resonance energy transfer nanoprobe

Chemical Communications ◽

10.1039/c8cc07211f ◽

2018 ◽

Vol 54 (88) ◽

pp. 12479-12482 ◽

Cited By ~ 17

Author(s):

Xueting Liu ◽

Nannan Fan ◽

Lijie Wu ◽

Chuanchen Wu ◽

Yongqing Zhou ◽

...

Keyword(s):

Alkaline Phosphatase ◽

Energy Transfer ◽

Liver Injury ◽

Alkaline Phosphatase Level ◽

Resonance Energy Transfer ◽

Resonance Energy ◽

Drug Induced ◽

Drug Induced Liver Injury

Ultra-sensitive imaging of the alkaline phosphatase levelin vivoin drug-induced liver injury with a new chemiluminescence resonance energy transfer nanoprobe.

Download Full-text

Ginsenoside Rb1, A Major Saponin from Panax ginseng, Exerts Protective Effects Against Acetaminophen-Induced Hepatotoxicity in Mice

The American Journal of Chinese Medicine ◽

10.1142/s0192415x19500927 ◽

2019 ◽

Vol 47 (08) ◽

pp. 1815-1831 ◽

Cited By ~ 4

Author(s):

Shen Ren ◽

Jing Leng ◽

Xing-Yue Xu ◽

Shuang Jiang ◽

Ying-Ping Wang ◽

...

Keyword(s):

Liver Injury ◽

Panax Ginseng ◽

Inflammatory Responses ◽

Interleukin 1 ◽

Protective Effects ◽

Ginsenoside Rb1 ◽

Drug Induced ◽

Drug Induced Liver Injury ◽

Oxide Synthase

Acute liver injury (ALI) induced by acetaminophen (APAP) is the main cause of drug-induced liver injury. Previous reports indicated liver failure could be alleviated by saponins (ginsenosides) from Panax ginseng against APAP-induced inflammatory responses in vivo. However, validation towards ginsenoside Rb1 as a major and marker saponin may protect liver from APAP-induced ALI and its mechanisms are poorly elucidated. In this study, the protective effects and the latent mechanisms of Rb1 action against APAP-induced hepatotoxicity were investigated. Rb1 was administered orally with 10[Formula: see text]mg/kg and 20[Formula: see text]mg/kg daily for 1 week before a single injection of APAP (250[Formula: see text]mg/kg, i.p.) 1[Formula: see text]h after the last treatment of Rb1. Serum alanine/aspartate aminotransferases (ALT/AST), liver glutathione (GSH) depletion, as well as the inflammatory cytokines, such as tumor necrosis factor-[Formula: see text] (TNF-[Formula: see text]), interleukin-1[Formula: see text] (IL-1[Formula: see text]), inducible nitric oxide synthase (iNOS), and cyclooxygenase-2 (COX-2), were analyzed to indicate the underlying protective effects of Rb1 against APAP-induced hepatotoxicity with significant inflammatory responses. Histological examination further proved Rb1’s protective effects. Importantly, Rb1 mitigated the changes in the phosphorylation of MAPK and PI3K/Akt, as well as its downstream factor NF-[Formula: see text]B. In conclusion, experimental data clearly demonstrated that Rb1 exhibited a remarkable liver protective effect against APAP-induced ALI, partly through regulating MAPK and PI3K/Akt signaling pathways-mediated inflammatory responses.

Download Full-text

In Silico Prediction of Drug-Induced Liver Injury Based on Ensemble Classifier Method

International Journal of Molecular Sciences ◽

10.3390/ijms20174106 ◽

2019 ◽

Vol 20 (17) ◽

pp. 4106 ◽

Cited By ~ 5

Author(s):

Wang ◽

Xiao ◽

Chen ◽

Wang

Keyword(s):

Liver Injury ◽

Drug Research ◽

Molecular Descriptors ◽

Ensemble Classifier ◽

Key Factors ◽

Drug Induced ◽

Molecular Fingerprints ◽

Injury Prediction ◽

Drug Induced Liver Injury ◽

Fold Cross Validation

Drug-induced liver injury (DILI) is a major factor in the development of drugs and the safety of drugs. If the DILI cannot be effectively predicted during the development of the drug, it will cause the drug to be withdrawn from markets. Therefore, DILI is crucial at the early stages of drug research. This work presents a 2-class ensemble classifier model for predicting DILI, with 2D molecular descriptors and fingerprints on a dataset of 450 compounds. The purpose of our study is to investigate which are the key molecular fingerprints that may cause DILI risk, and then to obtain a reliable ensemble model to predict DILI risk with these key factors. Experimental results suggested that 8 molecular fingerprints are very critical for predicting DILI, and also obtained the best ratio of molecular fingerprints to molecular descriptors. The result of the 5-fold cross-validation of the ensemble vote classifier method obtain an accuracy of 77.25%, and the accuracy of the test set was 81.67%. This model could be used for drug‐induced liver injury prediction.

Download Full-text

Zucker Lean Rats With Hepatic Steatosis Recapitulate Asymptomatic Metabolic Syndrome and Exhibit Greater Sensitivity to Drug-Induced Liver Injury Compared With Standard Nonclinical Sprague-Dawley Rat Model

Toxicologic Pathology ◽

10.1177/0192623320968716 ◽

2020 ◽

Vol 48 (8) ◽

pp. 994-1007

Author(s):

Timothy P. LaBranche ◽

Anna K. Kopec ◽

Srinivasa R. Mantena ◽

Brett D. Hollingshead ◽

Andrew W. Harrington ◽

...

Keyword(s):

Metabolic Syndrome ◽

Pharmaceutical Industry ◽

Liver Injury ◽

Hepatic Steatosis ◽

Sprague Dawley ◽

Drug Induced ◽

Drug Induced Liver Injury ◽

Sd Rats ◽

Normal Chow

Fatty liver disease is a potential risk factor for drug-induced liver injury (DILI). Despite advances in nonclinical in vitro and in vivo models to assess liver injury during drug development, the pharmaceutical industry is still plagued by idiosyncratic DILI. Here, we tested the hypothesis that certain features of asymptomatic metabolic syndrome (namely hepatic steatosis) increase the risk for DILI in certain phenotypes of the human population. Comparison of the Zucker Lean (ZL) and Zucker Fatty rats fed a high fat diet (HFD) revealed that HFD-fed ZL rats developed mild hepatic steatosis with compensatory hyperinsulinemia without increases in liver enzymes. We then challenged steatotic HFD-fed ZL rats and Sprague-Dawley (SD) rats fed normal chow, a nonclinical model widely used in the pharmaceutical industry, with acetaminophen overdose to induce liver injury. Observations in HFD-fed ZL rats included increased liver injury enzymes and greater incidence and severity of hepatic necrosis compared with similarly treated SD rats. The HFD-fed ZL rats also had disproportionately higher hepatic drug accumulation, which was linked with abnormal hepatocellular efflux transporter distribution. Here, we identify ZL rats with HFD-induced hepatic steatosis as a more sensitive nonclinical in vivo test system for modeling DILI compared with SD rats fed normal chow.

Download Full-text

Hepatotoxicity Modeling Using Counter-Propagation Artificial Neural Networks: Handling an Imbalanced Classification Problem

Molecules ◽

10.3390/molecules25030481 ◽

2020 ◽

Vol 25 (3) ◽

pp. 481 ◽

Cited By ~ 1

Author(s):

Benjamin Bajželj ◽

Viktor Drgan

Keyword(s):

Neural Networks ◽

Artificial Neural Networks ◽

Liver Injury ◽

In Vivo Studies ◽

Training Procedure ◽

Imbalanced Dataset ◽

Drug Induced ◽

Drug Induced Liver Injury ◽

Artificial Neural

Drug-induced liver injury is a major concern in the drug development process. Expensive and time-consuming in vitro and in vivo studies do not reflect the complexity of the phenomenon. Complementary to wet lab methods are in silico approaches, which present a cost-efficient method for toxicity prediction. The aim of our study was to explore the capabilities of counter-propagation artificial neural networks (CPANNs) for the classification of an imbalanced dataset related to idiosyncratic drug-induced liver injury and to develop a model for prediction of the hepatotoxic potential of drugs. Genetic algorithm optimization of CPANN models was used to build models for the classification of drugs into hepatotoxic and non-hepatotoxic class using molecular descriptors. For the classification of an imbalanced dataset, we modified the classical CPANN training algorithm by integrating random subsampling into the training procedure of CPANN to improve the classification ability of CPANN. According to the number of models accepted by internal validation and according to the prediction statistics on the external set, we concluded that using an imbalanced set with balanced subsampling in each learning epoch is a better approach compared to using a fixed balanced set in the case of the counter-propagation artificial neural network learning methodology.

Download Full-text

Current Perspective: 3D Spheroid Models Utilizing Human-Based Cells for Investigating Metabolism-Dependent Drug-Induced Liver Injury

Frontiers in Medical Technology ◽

10.3389/fmedt.2020.611913 ◽

2020 ◽

Vol 2 ◽

Author(s):

Christopher R. Cox ◽

Stephen Lynch ◽

Christopher Goldring ◽

Parveen Sharma

Keyword(s):

Liver Injury ◽

Pharmaceutical Companies ◽

Drug Induced ◽

Drug Induced Liver Injury ◽

Current Perspective ◽

Dosing Regimens ◽

Underlying Mechanisms ◽

Approved Drugs

Drug-induced liver injury (DILI) remains a leading cause for the withdrawal of approved drugs. This has significant financial implications for pharmaceutical companies, places increasing strain on global health services, and causes harm to patients. For these reasons, it is essential that in-vitro liver models are capable of detecting DILI-positive compounds and their underlying mechanisms, prior to their approval and administration to patients or volunteers in clinical trials. Metabolism-dependent DILI is an important mechanism of drug-induced toxicity, which often involves the CYP450 family of enzymes, and is associated with the production of a chemically reactive metabolite and/or inefficient removal and accumulation of potentially toxic compounds. Unfortunately, many of the traditional in-vitro liver models fall short of their in-vivo counterparts, failing to recapitulate the mature hepatocyte phenotype, becoming metabolically incompetent, and lacking the longevity to investigate and detect metabolism-dependent DILI and those associated with chronic and repeat dosing regimens. Nevertheless, evidence is gathering to indicate that growing cells in 3D formats can increase the complexity of these models, promoting a more mature-hepatocyte phenotype and increasing their longevity, in vitro. This review will discuss the use of 3D in vitro models, namely spheroids, organoids, and perfusion-based systems to establish suitable liver models to investigate metabolism-dependent DILI.

Download Full-text

TAMH: A Useful In Vitro Model for Assessing Hepatotoxic Mechanisms

BioMed Research International ◽

10.1155/2016/4780872 ◽

2016 ◽

Vol 2016 ◽

pp. 1-9 ◽

Cited By ~ 5

Author(s):

Madison Davis ◽

Brendan D. Stamper

Keyword(s):

Liver Injury ◽

Transforming Growth Factor ◽

Chemical Exposure ◽

Health Concern ◽

Hepatocellular Injury ◽

Drug Induced ◽

Endogenous Factors ◽

Drug Induced Liver Injury

In vitro models for hepatotoxicity can be useful tools to predict in vivo responses. In this review, we discuss the use of the transforming growth factor-αtransgenic mouse hepatocyte (TAMH) cell line, which is an attractive model to study drug-induced liver injury due to its ability to retain a stable phenotype and express drug-metabolizing enzymes. Hepatotoxicity involves damage to the liver and is often associated with chemical exposure. Since the liver is a major site for drug metabolism, drug-induced liver injury is a serious health concern for certain agents. At the molecular level, various mechanisms may protect or harm the liver during drug-induced hepatocellular injury including signaling pathways and endogenous factors (e.g., Bcl-2, GSH, Nrf2, or MAPK). The interplay between these and other pathways in the hepatocyte can change upon drug or drug metabolite exposure leading to intracellular stress and eventually cell death and liver injury. This review focuses on mechanistic studies investigating drug-induced toxicity in the TAMH line and how alterations to hepatotoxic mechanisms in this model relate to the in vivo situation. The agents discussed herein include acetaminophen (APAP), tetrafluoroethylcysteine (TFEC), flutamide, PD0325901, lapatinib, and flupirtine.

Download Full-text