Grantee Research Project Results
Final Report: Predictive QSAR Models of Hepatotoxicity
EPA Grant Number: R834999Title: Predictive QSAR Models of Hepatotoxicity
Investigators: Tropsha, Alex
Institution: University of North Carolina at Chapel Hill
EPA Project Officer: Chung, Serena
Project Period: June 1, 2011 through May 30, 2015
Project Amount: $750,000
RFA: Computational Toxicology: Biologically-Based Multi-Scale Modeling (2010) RFA Text | Recipients Lists
Research Category: Chemical Safety for Sustainability
Objective:
The objectives of this project were to (i) compile and curate a comprehensive, chemical hepatotoxicity database; (ii) develop rigorously validated and externally predictive models of chemical hepatotoxicity; and (iii) deploy curated data, computational model building tools, and hepatotoxicity models to public ChemBench web portal. In the course of the project, we have assembled a highly curated HepaToxDV database containing accurate structures hepatotoxicity annotations of chemicals, and developed several rigorous models of hepatotoxicity (Low et al, 2011, 2013). As a major outcome of this project, we have addressed a difficult issue of QSAR model interpretation and summarized our approaches to hybrid QSAR modeling in a major review.
Summary/Accomplishments (Outputs/Outcomes):
To comment on the latter issue, a significant part of model validation is its interpretation in terms of significant descriptors and most importantly, chemical features that can be understood by chemists. Although mechanistic interpretation of multivariate models is generally considered infeasible, it is still possible to look into weights of individual features (descriptors) and attempt to map features with significant weights into chemically sensible fragments, which results in the selection of some fragments as chemical toxicity alerts. The important fragments can be determined by analyzing the statistical significance of each chemical feature based on its contribution to the success of the model. However, we should be cognizant of the fact that individual features do not act independent of each other. Thus, we have developed a new methodology that looked at co-occurrence of significant features. This approach helped us develop new complex, integrated toxicity alerts that achieved unprecedented 100% accuracy in detecting toxic compounds (Low et al, 2015).
As another major accomplishment, we have developed an integrated computational framework that relies both on chemical descriptors and results of short term biological assays to improve the accuracy of toxicity prediction. In recent years, high throughput biological assays have been increasingly employed to elucidate mechanisms of chemical toxicity and predict toxic effects of chemicals in vivo. The data generated in such assays can be considered as biological descriptors of chemicals that can be combined with molecular descriptors and employed in QSAR modeling to improve the accuracy of toxicity prediction. In a recent review (Low et al, 2014), we have discussed several approaches for integrating chemical and biological data for predicting biological effects of chemicals in vivo and compare their performance across several data sets.
Conclusions:
We concluded that while no method consistently shows superior performance, the integrative approaches rank consistently among the best yet offer enriched interpretation of models over those built with either chemical or biological data alone.
Journal Articles on this Report : 4 Displayed | Download in RIS Format
| Other project views: | All 4 publications | 4 publications in selected types | All 4 journal articles |
|---|
| Type | Citation | ||
|---|---|---|---|
|
|
Low YS, Sedykh AY, Rusyn I, Tropsha A. Integrative approaches for predicting in vivo effects of chemicals from their structural descriptors and the results of short-term biological assays. Current Topics in Medicinal Chemistry 2014;14(11):1356-1364. |
R834999 (Final) R832720 (2009) R833825 (Final) R835166 (2014) R835166 (2016) R835166 (Final) |
Exit |
|
|
Low Y, Uehara T, Minowa Y, Yamada H, Ohno Y, Urushidani T, Sedykh A, Muratov E, Kuz’min V, Fourches D, Zhu H, Rusyn I, Tropsha A. Predicting drug-induced hepatotoxicity using QSAR and toxicogenomics approaches. Chemical Research in Toxicology 2011;24(8):1251-1262. |
R834999 (Final) R833825 (Final) |
Exit Exit Exit |
|
|
Low Y, Sedykh A, Fourches D, Golbraikh A, Whelan M, Rusyn I, Tropsha A. Integrative chemical-biological read-across approach for chemical hazard classification. Chemical Research in Toxicology 2013;26(8):1199-1208. |
R834999 (Final) R835166 (2013) R835166 (2016) R835166 (Final) |
Exit Exit Exit |
|
|
Low YS, Caster O, Bergvall T, Fourches D, Zang X, Norén GN, Rusyn I, Edwards R, Tropsha A. Cheminformatics-aided pharmacovigilance:application to Stevens-Johnson Syndrome. Journal of the American Medical Informatics Association 2016;23(5):968-978. |
R834999 (Final) |
|
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.