Innovative Strategies to Develop Chemical Categories Using a Combination of Structural and Toxicological Properties

Abstract

Interest is increasing in the development of non-animal methods for toxicological evaluations. These methods are however, particularly challenging for complex toxicological endpoints such as repeated dose toxicity. European Legislation, e.g., the European Union's Cosmetic Directive and REACH, demands the use of alternative methods. Frameworks, such as the Read-across Assessment Framework or the Adverse Outcome Pathway Knowledge Base, support the development of these methods. The aim of the project presented in this publication was to develop substance categories for a read-across with complex endpoints of toxicity based on existing databases. The basic conceptual approach was to combine structural similarity with shared mechanisms of action. Substances with similar chemical structure and toxicological profile form candidate categories suitable for read-across. We combined two databases on repeated dose toxicity, RepDose database, and ELINCS database to form a common database for the identification of categories. The resulting database contained physicochemical, structural, and toxicological data, which were refined and curated for cluster analyses. We applied the Predictive Clustering Tree (PCT) approach for clustering chemicals based on structural and on toxicological information to detect groups of chemicals with similar toxic profiles and pathways/mechanisms of toxicity. As many of the experimental toxicity values were not available, this data was imputed by predicting them with a multi-label classification method, prior to clustering. The clustering results were evaluated by assessing chemical and toxicological similarities with the aim of identifying clusters with a concordance between structural information and toxicity profiles/mechanisms. From these chosen clusters, seven were selected for a quantitative read-across, based on a small ratio of NOAEL of the members with the highest and the lowest NOAEL in the cluster (< 5). We discuss the limitations of the approach. Based on this analysis we propose improvements for a follow-up approach, such as incorporation of metabolic information and more detailed mechanistic information. The software enables the user to allocate a substance in a cluster and to use this information for a possible read- across. The clustering tool is provided as a free web service, accessible at http://mlc-reach.informatik.uni-mainz.de.

In silicotoxicology: comprehensive benchmarking of multi‐label classification methods applied to chemical toxicity data

“…Predicting continuous endpoints can eliminate the discretization step of toxicity data, which requires endpoint-specific olds. 23 However, it was shown in another study that predicting exact values of continuous endpoints is harder than predicting binary categories. 14 Some multi-label classification methods, such as Random K labelset and label powerset, are not suitable for continuous endpoints.…”

“…Developments have been made to create multi-label classification models using toxicity data sets such as the ToxCast data set, 15,16 Tox21 data sets, [17][18][19] Accelrys Toxicity Database, 20,21 and RepDose and ELINC data sets. 22,23 However, a recurrent problem in these studies, caused by missing labels in toxicity data sets, illustrated a major challenge in applying multi-label classification approaches to toxicity prediction. Toxicity profiles of some compounds are unknown across all toxicity phenotypes, either because such data are unavailable (e.g., compounds are not tested for all toxicity phenotypes), or it may be hard to find (e.g., being scattered in scientific literature).…”

“…Additionally, some studies applied imputation as a preprocessing step. 8,[22][23][24] Imputation is a process that aims at 'completing' the missing data in the data set by substituting (in this case) the missing toxicity data with predicted values. Subsequently, multi-label classification methods are applied to the completed data set.…”

In silico prediction of dermal absorption of pesticides – an evaluation of selected models against results from in vitro testing

“…Correlation analysis revealed that the predictivity was not influenced by the molecular weight of the active substances in the examined dataset. From a physiological point of view, dermal absorption is thought to decrease in particular when compounds larger than 500 g/mol are presented to the skin [34,35]. The examined dataset is restricted regarding the molecular weight range with only three compounds exceeding 500 g/mol.…”

Lessons learned from read-across case studies for repeated-dose toxicity

“…Since the review of Cronin et al (2013), a number of papers have appeared that focus on modern-day RA. Many of these, including Blackburn and Stuard (2014), European Chemicals Agency (ECHA) (2015), Organisation for Economic Co-operation and Development (OECD) (2015) and Schultz et al (2015), have put forward efforts to improve RA arguments and improve and innovate approaches (Batke et al, 2016;de Abrew et al, 2016;Shah et al, 2016;van Ravenzwaay et al, 2016). More recently, Ball et al, (2016) summarised the state-of-the-art surrounding read-across, along with reasons relating to regulatory non-acceptance, and compiled relevant guidance under the heading of "Good Read-Across Practice"; Hartung (2016) described the concept of linking different types of data and tools under the umbrella of good read-across practices.…”