Predictive Model for Drug-Induced Liver Injury Using Deep Neural Networks Based on Substructure Space

Drug-induced liver injury (DILI) is a major concern for drug developers, regulators, and clinicians. However, there is no adequate model system to assess drug-associated DILI risk in humans. In the big data era, computational models are expected to play a revolutionary role in this field. This study aimed to develop a deep neural network (DNN)-based model using extended connectivity fingerprints of diameter 4 (ECFP4) to predict DILI risk. Each data set for the predictive model was retrieved and curated from DILIrank, LiverTox, and other literature. The best model was constructed through ten iterations of stratified 10-fold cross-validation, and the applicability domain was defined based on integer ECFP4 bits of the training set which represented substructures. For the robustness test, we employed the concept of the endurance level. The best model showed an accuracy of 0.731, a sensitivity of 0.714, and a specificity of 0.750 on the validation data set in the complete applicability domain. The model was further evaluated with four external data sets and attained an accuracy of 0.867 on 15 drugs with DILI cases reported since 2019. Overall, the results suggested that the ECFP4-based DNN model represents a new tool to identify DILI risk for the evaluation of drug safety.

Download Full-text

Computational Models Using Multiple Machine Learning Algorithms for Predicting Drug Hepatotoxicity with the DILIrank Dataset

10.20944/preprints202002.0178.v1 ◽

2020 ◽

Author(s):

Robert Ancuceanu ◽

Marilena Viorica Hovanet ◽

Adriana Iuliana Anghel ◽

Florentina Furtunescu ◽

Monica Neagu ◽

...

Keyword(s):

Machine Learning ◽

Liver Injury ◽

Computational Models ◽

Liver Toxicity ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Drug Induced ◽

Reference Drug ◽

Drug Induced Liver Injury

Drug induced liver injury (DILI) remains one of the challenges in the safety profile of both authorized drugs and candidate drugs and predicting hepatotoxicity from the chemical structure of a substance remains a challenge worth pursuing, being also coherent with the current tendency for replacing non-clinical tests with in vitro or in silico alternatives. In 2016 a group of researchers from FDA published an improved annotated list of drugs with respect to their DILI risk, constituting “the largest reference drug list ranked by the risk for developing drug-induced liver injury in humans”, DILIrank. This paper is one of the few attempting to predict liver toxicity using the DILIrank dataset. Molecular descriptors were computed with the Dragon 7.0 software, and a variety of feature selection and machine learning algorithms were implemented in the R computing environment. Nested (double) cross-validation was used to externally validate the models selected. A number of 78 models with reasonable performance have been selected and stacked through several approaches, including the building of multiple meta-models. The performance of the stacked models was slightly superior to other models published. The models were applied in a virtual screening exercise on over 100,000 compounds from the ZINC database and about 20% of them were predicted to be non-hepatotoxic.

Download Full-text

Computational Models Using Multiple Machine Learning Algorithms for Predicting Drug Hepatotoxicity with the DILIrank Dataset

International Journal of Molecular Sciences ◽

10.3390/ijms21062114 ◽

2020 ◽

Vol 21 (6) ◽

pp. 2114

Author(s):

Robert Ancuceanu ◽

Marilena Viorica Hovanet ◽

Adriana Iuliana Anghel ◽

Florentina Furtunescu ◽

Monica Neagu ◽

...

Keyword(s):

Machine Learning ◽

Liver Injury ◽

Computational Models ◽

Liver Toxicity ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Drug Induced ◽

Reference Drug ◽

Drug Induced Liver Injury

Drug-induced liver injury (DILI) remains one of the challenges in the safety profile of both authorized and candidate drugs, and predicting hepatotoxicity from the chemical structure of a substance remains a task worth pursuing. Such an approach is coherent with the current tendency for replacing non-clinical tests with in vitro or in silico alternatives. In 2016, a group of researchers from the FDA published an improved annotated list of drugs with respect to their DILI risk, constituting “the largest reference drug list ranked by the risk for developing drug-induced liver injury in humans” (DILIrank). This paper is one of the few attempting to predict liver toxicity using the DILIrank dataset. Molecular descriptors were computed with the Dragon 7.0 software, and a variety of feature selection and machine learning algorithms were implemented in the R computing environment. Nested (double) cross-validation was used to externally validate the models selected. A total of 78 models with reasonable performance were selected and stacked through several approaches, including the building of multiple meta-models. The performance of the stacked models was slightly superior to other models published. The models were applied in a virtual screening exercise on over 100,000 compounds from the ZINC database and about 20% of them were predicted to be non-hepatotoxic.

Download Full-text

Stem cell-derived hepatocytes as a predictive model for drug-induced liver injury: are we there yet?

British Journal of Clinical Pharmacology ◽

10.1111/j.1365-2125.2012.04360.x ◽

2013 ◽

Vol 75 (4) ◽

pp. 885-896 ◽

Cited By ~ 56

Author(s):

Richard Kia ◽

Rowena L. C. Sison ◽

James Heslop ◽

Neil R. Kitteringham ◽

Neil Hanley ◽

...

Keyword(s):

Stem Cell ◽

Liver Injury ◽

Predictive Model ◽

Drug Induced ◽

Drug Induced Liver Injury

Download Full-text

DRUG-INDUCED LIVER INJURY: A PREDICTIVE MODEL, MITOCHONDRIAL TOXICITY MECHANISMS RISK, AND ANTIPSYCHOTIC MEDICATIONS RISK IN A REAL-WORLD SETTING

10.23860/rana-payal-2020 ◽

2020 ◽

Author(s):

◽

Payal Rana

Keyword(s):

Liver Injury ◽

Predictive Model ◽

Real World ◽

Mitochondrial Toxicity ◽

Drug Induced ◽

Antipsychotic Medications ◽

Drug Induced Liver Injury ◽

Real World Setting ◽

Toxicity Mechanisms

Download Full-text

A compound attributes-based predictive model for drug induced liver injury in humans

PLoS ONE ◽

10.1371/journal.pone.0231252 ◽

2020 ◽

Vol 15 (4) ◽

pp. e0231252

Author(s):

Yang Liu ◽

Hua Gao ◽

Yudong D. He

Keyword(s):

Liver Injury ◽

Predictive Model ◽

Drug Induced ◽

Drug Induced Liver Injury

Download Full-text

Association Between the Characteristics of Small-molecule Kinase Inhibitors and Drug-induced Liver Injury: A Scoping Study (Preprint)

10.2196/preprints.33257 ◽

2021 ◽

Author(s):

Huiqun Dong ◽

Jia You ◽

Yu Zhao ◽

Danhua Zheng ◽

Yi Zhong ◽

...

Keyword(s):

Liver Injury ◽

Small Molecule ◽

Odds Ratio ◽

Kinase Inhibitors ◽

Mitochondrial Proteins ◽

Binding Potential ◽

Defined Daily Dose ◽

Drug Induced ◽

Data Set ◽

Drug Induced Liver Injury

BACKGROUND Preventing Drug-induced liver injury (DILI) in advance is an important task to improve drug safety and protect patient health. It was reported that more than half of small-molecule kinase inhibitors (KIs) induced DILI clinically. Meanwhile, numerous studies have shown a close relationship between mitochondrial damage and the generation of DILI. OBJECTIVE We aimed to focused on KIs to find factors related to DILI occurrence and study the binding potential between the whole class of drugs and mitochondrial proteins and further analysis of the key proteins in silico. METHODS 1,223 drugs approved by the Food and Drug Administration (FDA) were collected and analyzed, including 44 KIs. Fisher exact test was used to analyze DILI potential and risk of different factors. 187 human mitochondrial proteins were further collected and high-throughput molecular docking was performed between them and drugs in the data set. RESULTS Be The possibility of KIs to produce DILI is much higher than other types (odds ratio [OR] = 46.89, 95% CI [confidence interval] = 6.44 ~ 341.63, P = 9.28E-13). A few DILI risk factors were found, including molecular weight (MW) between 400 and 600, the defined daily dose (DDD) greater than or equal to 100mg/day, the octanol-water partition coefficient (LogP) greater than or equal to 3, and the degree of liver metabolism (LM) more than 50% (‘400 ≤ MW < 600 & LogP ≥ 3 & DDD ≥ 100 & LM ≥ 50%’). Drugs that met this combination of rules were found to have higher DILI risk than controls (OR = 8.28, 95% CI = 2.46 ~ 27.82, P = 4.82E-05, PPV [positive predictive value] = 88%) and were more likely to cause severe DILI (OR = 8.26, 95% CI = 2.25 ~ 30.26, P = 5.06E-04). The docking results showed that KIs had significant higher affinity with human mitochondrial proteins (P = 4.19E-11), which may be an implication for higher DILI potential. CONCLUSIONS KIs were found to have the highest odds ratio of causing DILI. Some characteristics of KIs were significantly related to the production of DILI. And the average docking scores of KIs drugs were found to be significant different from other classes. Further analysis identified the top binding mitochondrial proteins for KIs, which may help with the study of the mechanism of DILI. CLINICALTRIAL Not applicable.

Download Full-text

Machine-Learning Prediction of Oral Drug-Induced Liver Injury (DILI) via Multiple Features and Endpoints

BioMed Research International ◽

10.1155/2020/4795140 ◽

2020 ◽

Vol 2020 ◽

pp. 1-10

Author(s):

Xiaobin Liu ◽

Danhua Zheng ◽

Yi Zhong ◽

Zhaofan Xia ◽

Heng Luo ◽

...

Keyword(s):

Machine Learning ◽

Drug Discovery ◽

Liver Injury ◽

Computational Models ◽

Characteristic Curve ◽

Prediction Performance ◽

Oral Drug ◽

Good Prediction ◽

Drug Induced ◽

Drug Induced Liver Injury

Drug discovery is a costly process which usually takes more than 10 years and billions of dollars for one successful drug to enter the market. Despite all the safety tests, drugs may still cause adverse reactions and be restricted in use or even withdrawn from the market. Drug-induced liver injury (DILI) is one of the major adverse drug reactions, and computational models may be used to predict and reduce it. To assess the computational prediction performance of DILI, we curated DILI endpoints from three databases and prepared drug features including chemical descriptors, therapeutic classifications, gene expressions, and binding proteins. We trained machine-learning models to predict the various DILI endpoints using different drug features. Using the optimal feature sets, the top-performing models obtained areas under the receiver operating characteristic curve (AUC) around 0.8 for some DILI endpoints. We found that some features, including therapeutic classifications and proteins, have good prediction performance towards DILI. We also discovered that the severity of DILI endpoints as well as the selection of negative samples may significantly affect the prediction results. Overall, our study provided a comprehensive collection, curation, and prediction of DILI endpoints using various drug features, which may help the drug researchers to better understand and prevent DILI during the drug discovery process.

Download Full-text

Leberschaden unter oraler Antikoagulation, Statin- und ACE-Hemmertherapie

Praxis ◽

10.1024/1661-8157/a000298 ◽

2010 ◽

Vol 99 (21) ◽

pp. 1259-1265 ◽

Cited By ~ 2

Author(s):

Bruggisser ◽

Terraciano ◽

Rätz Bravo ◽

Haschke

Keyword(s):

Liver Injury ◽

Wird Eine ◽

Drug Induced ◽

Drug Induced Liver Injury

Ein 71-jähriger Patient stellt sich mit Epistaxis und ikterischen Skleren auf der Notfallstation vor. Der Patient steht unter einer Therapie mit Phenprocoumon, Atorvastatin und Perindopril. Anamnestisch besteht ein langjähriger Alkoholabusus. Laborchemisch werden massiv erhöhte Leberwerte (ALAT, Bilirubin) gesehen. Der INR ist unter oraler Antikoagulation und bei akuter Leberinsuffizienz >12. Die weiterführenden Abklärungen schliessen eine Virushepatitis und eine Autoimmunhepatitis aus. Nachdem eine Leberbiopsie durchgeführt werden kann, wird eine medikamentös-toxische Hepatitis, ausgelöst durch die Komedikation von Atorvastatin, Phenprocoumon und Perindopril bei durch Alkohol bereits vorgeschädigter Leber diagnostiziert. Epidemiologie, Pathophysiologie und Klink der medikamentös induzierten Leberschäden (drug induced liver injury, DILI), speziell von Coumarinen, Statinen und ACE-Hemmern werden im Anschluss an den Fallbericht diskutiert.

Download Full-text