Search

The lazar framework for read across predictions was expanded for the prediction of nanoparticle toxicities, and a new methodology for calculating nanoparticle descriptors from core and coating structures was implemented. Nano-lazar provides a flexible and reproducible framework for downloading data and ontologies from the open eNanoMapper infrastructure, developing and validating nanoparticle read across models, open-source code and a free graphical interface for nanoparticle read-across predictions. In this study we compare different nanoparticle descriptor sets and local regression algorithms. Sixty independent crossvalidation experiments were performed for the Net Cell Association endpoint of the Protein Corona dataset. The best RMSE and r2 results originated from models with protein corona descriptors and the weighted random forest algorithm, but their 95% prediction interval is significantly less accurate than for models with simpler descriptor sets (measured and calculated nanoparticle properties). The most accurate prediction intervals were obtained with measured nanoparticle properties (no statistical significant difference (p < 0.05) of RMSE and r2 values compared to protein corona descriptors). Calculated descriptors are interesting for cheap and fast high-throughput screening purposes. RMSE and prediction intervals of random forest models are comparable to protein corona models, but r2 values are significantly lower.

The aim of the SEURAT-1 (Safety Evaluation Ultimately Replacing Animal Testing-1) research cluster, comprised of seven EU FP7 Health projects co-financed by Cosmetics Europe, is to generate a proof-of-concept to show how the latest technologies, systems toxicology and toxicogenomics can be combined to deliver a test replacement for repeated dose systemic toxicity testing on animals. The SEURAT-1 strategy is to adopt a mode-of-action framework to describe repeated dose toxicity, combining in vitro and in silico methods to derive predictions of in vivo toxicity responses. ToxBank is the cross-cluster infrastructure project whose activities include the development of a data warehouse to provide a web-accessible shared repository of research data and protocols, a physical compounds repository, reference or 'gold compounds' for use across the cluster (available via wiki.toxbank.net), and a reference resource for biomaterials. Core technologies used in the data warehouse include the ISA-Tab universal data exchange format, REpresentational State Transfer (REST) web services, the W3C Resource Description Framework (RDF) and the OpenTox standards. We describe the design of the data warehouse based on cluster requirements, the implementation based on open standards, and finally the underlying concepts and initial results of a data analysis utilizing public data related to the gold compounds.

Deliverable 5.7 reports on the Tasks 5.3 and 5.9 on the user applications achievements. The application infrastructure developed within eNanoMapper project aims to support the data management in the area of nanosafety research and to enable an integrated approach for the risk assessment of nanomaterials. To achieve these, eNanoMapper developed an ontology, a data infrastructure and modelling tools with applicability in risk assessment of nanomaterials. eNanoMapper developed resources and tools for predicting toxicity of nanomaterials and worked towards improving the standards in risk assessment of nanomaterials.

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. peerReviewed

Intestinal drug absorption in humans is a central topic in drug discovery. In this study, we use a broad selection of machine learning and statistical methods for the classification and numerical prediction of this key end point. Our data set is based on a selection of 458 small druglike compounds with FDA approval. Using easily available tools, we calculated one- to three-dimensional physicochemical descriptors and used various methods of feature selection (best-first backward selection, correlation analysis, and decision tree analysis). We then used decision tree induction (DTI), fragment-based lazy-learning (LAZAR), support vector machine classification, multilayer perceptrons, random forests, k-nearest neighbor and Naïve Bayes analysis to model absorption ratios and binary classification (well-absorbed and poorly absorbed compounds). Best performance for classification was seen with DTI using the chi-squared analysis interaction detector (CHAID) algorithm, yielding corrected classification rate of 88% (Matthews correlation coefficient of 75%). In numeric predictions, the multilayer perceptron performed best, achieving a root mean squared error of 25.823 and a coefficient of determination of 0.6. In line with current understanding is the importance of descriptors such as lipophilic partition coefficients (log P) and hydrogen bonding. However, we are able to highlight the utility of gravitational indices and moments of inertia, reflecting the role of structural symmetry in oral absorption. Our models are based on a diverse data set of marketed drugs representing a broad chemical space. These models therefore contribute substantially to the molecular understanding of human intestinal drug absorption and qualify for a generalized use in drug discovery and lead optimization.

Drug safety is of great importance to public health. The detrimental effects of drugs not only limit their application but also cause suffering in individual patients and evoke distrust of pharmacotherapy. For the purpose of identifying drugs that could be suspected of causing adverse reactions, we present a structure-activity relationship analysis of adverse drug reactions (ADRs) in the central nervous system (CNS), liver, and kidney, and also of allergic reactions, for a broad variety of drugs (n = 507) from the Swiss drug registry. Using decision tree induction, a machine learning method, we determined the chemical, physical, and structural properties of compounds that predispose them to causing ADRs. The models had high predictive accuracies (78.9-90.2%) for allergic, renal, CNS, and hepatic ADRs. We show the feasibility of predicting complex end-organ effects using simple models that involve no expensive computations and that can be used (i) in the selection of the compound during the drug discovery stage, (ii) to understand how drugs interact with the target organ systems, and (iii) for generating alerts in postmarketing drug surveillance and pharmacovigilance.

Mutagenicity and carcinogenicity are endpoints of major environmental and regulatory concern. These endpoints are also important targets for development of alternative methods for screening and prediction due to the large number of chemicals of potential concern and the tremendous cost (in time, money, animals) of rodent carcinogenicity bioassays. Both mutagenicity and carcinogenicity involve complex, cellular processes that are only partially understood. Advances in technologies and generation of new data will permit a much deeper understanding. In silico methods for predicting mutagenicity and rodent carcinogenicity based on chemical structural features, along with current mutagenicity and carcinogenicity data sets, have performed well for local prediction (i.e., within specific chemical classes), but are less successful for global prediction (i.e., for a broad range of chemicals). The predictivity of in silico methods can be improved by improving the quality of the data base and endpoints used for modelling. In particular, in vitro assays for clastogenicity need to be improved to reduce false positives (relative to rodent carcinogenicity) and to detect compounds that do not interact directly with DNA or have epigenetic activities. New assays emerging to complement or replace some of the standard assays include Vitotox, GreenScreenGC, and RadarScreen. The needs of industry and regulators to assess thousands of compounds necessitate the development of high-throughput assays combined with innovative data-mining and in silico methods. Various initiatives in this regard have begun, including CAESAR, OSIRIS, CHEMOMENTUM, CHEMPREDICT, OpenTox, EPAA, and ToxCast. In silico methods can be used for priority setting, mechanistic studies, and to estimate potency. Ultimately, such efforts should lead to improvements in application of in silico methods for predicting carcinogenicity to assist industry and regulators and to enhance protection of public health.

eNanoMapper developed a modular infrastructure for data storage, sharing and searching, an ontologyfor the categorisation and characterisation of nanomaterials andcomputational models fornanomaterials safety assessment.

Several qualitative (hazard-based) models for chronic toxicity prediction are available through commercial and freely available software, but in the context of risk assessment a quantitative value is mandatory in order to be able to apply a Margin of Exposure (predicted toxicity/exposure estimate) approach to interpret the data. Recently quantitative models for the prediction of the carcinogenic potency have been developed, opening some hopes in this area, but this promising approach is currently limited by the fact that the proposed programs are neither publically nor commercially available. In this article we describe how two models (one for mouse and one for rat) for the carcinogenic potency (TD50) prediction have been developed, using lazar (Lazy Structure Activity Relationships), a procedure similar to read-across, but automated and reproducible. The models obtained have been compared with the recently published ones, resulting in a similar performance. Our aim is also to make the models freely available in the near future thought a user friendly internet web site.

We compared 121 replicate rodent carcinogenicity assays from the two parts (National Cancer Institute/National Toxicology Program and literature) of the Carcinogenic Potency Database (CPDB) to estimate the reliability of these experiments. We estimated a concordance of 57% between the overall rodent carcinogenicity classifications from both sources. This value did not improve substantially when additional biologic information (species, sex, strain, target organs) was considered. These results indicate that rodent carcinogenicity assays are much less reproducible than previously expected, an effect that should be considered in the development of structure-activity relationship models and the risk assessment process.

Every technique for toxicity prediction and for the detection of structure-activity relationships relies on the accurate estimation and representation of chemical and toxicologic properties. In this paper we discuss the potential sources of errors associated with the identification of compounds, the representation of their structures, and the calculation of chemical descriptors. It is based on a case study where machine learning techniques were applied to data from noncongeneric compounds and a complex toxicologic end point (carcinogenicity). We propose methods applicable to the routine quality control of large chemical datasets, but our main intention is to raise awareness about this topic and to open a discussion about quality assurance in predictive toxicology. The accuracy and reproducibility of toxicity data will be reported in another paper.

The single-cell gel electrophoresis (or comet) assay has gained widespread acceptance as a cheap and simple genotoxicity test, but it requires a computer-assisted image-analysis system. As commercial programs are expensive and inflexible, we decided to develop an image-analysis system based on public domain programs and make it publicly available for the scientific community. Our system is based on the scientific image-processing program NIH Image, and was written in its Pascal-like macro language. User interaction was kept as simple as possible, to enable the measurement of a large number of cells with a few keystrokes. Therefore, the time for image analysis is very low, even on slow computers. The comet macro can be obtained from http://mailbox.univie.ac.at/christoph.helma++ +/comet/, NIH Image is available at http://rsb.info.nih.gov/nih-image/. Both programs are free of charge.

The primary aim of the present investigation was to study possible adverse effects of groundwater from an aquifer south of Austria's capital, Vienna, and to relate these toxicological effects to routinely measured physical/chemical parameters. Fourty-three water samples were tested for genotoxic and ecotoxic effects. For genotoxicity testing the Salmonella/microsome assay, the micronucleus test with primary rat hepatocytes and micronucleus tests with plants( Tradescantia, Vicia faba) were used. In ecotoxicity tests, algae (Selenastrum capricornutum), water cress (Lepidium sativum), and water flea (Dapnia magna) were studied as target organisms. In genotoxicity assays, 10 samples (23%) gave a weak positive response with a single end point, but only one sample (2%) was genotoxic in three different test systems. Thirty-six samples (86%) caused adverse effects in ecotoxicity assays. Plants(algae and water cress) were more sensitive than daphnie. No correlations between toxic effects and physical/chemical parameters were detected. The genotoxicity experiments indicate presently a low risk from genotoxic compounds. The ecotoxic (especially phytotoxic) properties of many water samples raise concern about their suitability for irrigation purposes. The lacking correlation between results from toxicity tests and physical/chemical data indicates that it is presently impossible to predict toxic properties from routine physical/chemical measurements with a sufficient level of safety. It is therefore important to include biological toxicity assays in groundwater monitoring programs.

Neben Abwassern aus der Industrie entstehen auch bei anderen technischen Prozessen Abwasser, die gentoxische Substanzen enthalten konnen. Dabei handelt es sich sowohl um punktformige Quellen wie z.B. dem Bergbau, Kraftwerken oder Anlagen zur Abfall- bzw. Abwasserentsorgung als auch um diffuse Emissionen aus Landwirtschaft und Verkehr. Obwohl eine Vielzahl von mutagenen Verbindungen bekannt ist, war es in konkreten Fallen noch nicht moglich, die mutagene Aktivitat von Wassern bestimmten Substanzen zuzuordnen. Bei der Chlorierung und Ozonierung im Rahmen der Trinkwasseraufbereitung konnen ebenfalls Stoffe mit mutagenem Potential entstehen. Dieses Gesundheitsrisiko fur den menschen kann derzeit noch nicht abschliesend beurteilt werden.

The aim of this study was to evaluate the perfomance of four bacterial short-term genotoxicity assays (Salmonella/microsome assay, SOS Chromotest, Microscreen phage-induction assay, differential DNA repair test) that are widely used and/or have a promising potential for the genotoxicity testing of water samples. Twenty-three samples of different origins (drinking and bathing water, surface water, municipal and industrial wastewater, pulp mill effluents, groundwater, and landfill leachates) were tested in these assays. In total, 20 samples were genotoxic: 13 in the Salmonella/microsome assay, 13 in the SOS Chromotest, 8 in the Microscreen phage-induction assay, and 19 in the differential DNA repair test. Although the differential DNA repair test was the most sensitive system, positive results were obtained also with some of the negative control samples, and it had the least power to detect different genotoxic potencies. The Microscreen assay was the least sensitive system due to nonlinear results and samp...

This is the first report of clastogenic effects of chlorinated hydroxyfuranones (CHFs) in plants. Two byproducts of water chlorination, 3-chloro-4-(dichloromethyl)-5-hydroxy-2[5H]-furanone (MX) and 3,4-dichloro-5-hydroxy-2[5H]-furanone (MA) induced a dose dependent increase of micronuclei (MN) in pollen mother cells of Tradescantia when doses up to 100 micrograms MX and 500 micrograms MA were applied directly to the inflorescences. In contrast, exposure of the stems in aqueous solutions containing up to 1 mg/l MX and 10 mg/l MA did not cause a positive response.

(Q)SAR model validation is essential to ensure the quality of inferred models and to indicate future model predictivity on unseen compounds. Proper validation is also one of the requirements of regulatory authorities in order to accept the (Q)SAR model, and to approve its use in real world scenarios as alternative testing method. However, at the same time, the question of how to validate a (Q)SAR model, in particular whether to employ variants of cross-validation or external test set validation, is still under discussion. In this paper, we empirically compare a k-fold cross-validation with external test set validation. To this end we introduce a workflow allowing to realistically simulate the common problem setting of building predictive models for relatively small datasets. The workflow allows to apply the built and validated models on large amounts of unseen data, and to compare the performance of the different validation approaches. The experimental results indicate that cross-validation produces higher performant (Q)SAR models than external test set validation, reduces the variance of the results, while at the same time underestimates the performance on unseen compounds. The experimental results reported in this paper suggest that, contrary to current conception in the community, cross-validation may play a significant role in evaluating the predictivity of (Q)SAR models.

Die Kontamination von Gewassern mit gentoxischen Substanzen ist sowohl aus human- als auch aus okotoxikologischen Gesichtspunkten bedenklich. In epidemiologischen Studien wurden Korrelationen zwischen gentoxischen Gewasserbelastungen und kanzerogenen Effekten beim Menschen aber auch bei Fischen und Muscheln hergestellt. In dem vorliegenden Ubersichtsartikel werden biologische Methoden vorgestellt, die die Detektion gentoxischer Substanzen in Gewassern ermoglichen. Gegenuber chemisch-analytischen Methoden weisen sie den Vorteil auf, das nicht Einzelsubstanzen, sondern die Kombinationswirkung aller Schadstoffe in einem biologischen Testsystem erfast werden. Zur Prufung der Gentoxizitat kommen sowohlin vitro als auchin vivo Systeme zum Einsatz. Als Indikatororganismen dienen Bakterien, Pflanzen und Tiere, bei denen primare DNA-Schaden, Punkt-, Chromosomen- und Genommutationen als gentoxische Endpunkte nachgewiesen werden. Die Vor- und Nachteile der verschiedenen Testsysteme, ihre Einsatzmoglichkeit und die Kombination mehrerer Verfahren zu Testbatterien werden diskutiert und Methoden zur Konzentrierung von Wasserinhaltsstoffen vorgestellt.

The Tradescantia micronucleus (Trad-MCN) assay was used to determine clastogenic effects of contaminated groundwater collected near a hazardous waste landfill. Water samples were taken from a purification plant (activated charcoal filtration, UV irradiation) which was built to avoid groundwater contamination by this landfill. Five series of experiments were conducted during approximately 4 months. In addition, water samples were irradiated under laboratory conditions with increasing doses of UV light. Several field water samples gave positive, dose-dependent effects before filtration and irradiation. Maximal values ( 6.1 ± 4.7 micronuclei (MCN)/100 tetrads) were six-fold above controls. UV irradiation of activated charcoal-filtered water resulted in an enhancement of MCN frequencies. Exposure of groundwater to UV irradiation in the laboratory led to a dose-dependent increase of micronuclei. At the highest dose (1500 J/2) the MCN frequency was more than six times higher than in the unirradiated sample ( 5.4 ± 1.0 vs. 0.8 ± 0.4 MCN/100 tetrads). The clastogenicity of UV-irradiated samples decreased with a half-life of approximately 1 day. Irradiation of tap water did not increase the MCN frequency. Our results indicate that irradiation of water with UV light for disinfection purposes might lead to a transiently increased genotoxicity of chemically polluted water samples.

OpenTox provides an interoperable, standards-based Framework for the support of predictive toxicology data management, algorithms, modelling, validation and reporting. It is relevant to satisfying the chemical safety assessment requirements of the REACH legislation as it supports access to experimental data, (Quantitative) Structure-Activity Relationship models, and toxicological information through an integrating platform that adheres to regulatory requirements and OECD validation principles. Initial research defined the essential components of the Framework including the approach to data access, schema and management, use of controlled vocabularies and ontologies, architecture, web service and communications protocols, and selection and integration of algorithms for predictive modelling. OpenTox provides end-user oriented tools to non-computational specialists, risk assessors, and toxicological experts in addition to Application Programming Interfaces (APIs) for developers of new applications. OpenTox actively supports public standards for data representation, interfaces, vocabularies and ontologies, Open Source approaches to core platform components, and community-based collaboration approaches, so as to progress system interoperability goals. The OpenTox Framework includes APIs and services for compounds, datasets, features, algorithms, models, ontologies, tasks, validation, and reporting which may be combined into multiple applications satisfying a variety of different user needs. OpenTox applications are based on a set of distributed, interoperable OpenTox API-compliant REST web services. The OpenTox approach to ontology allows for efficient mapping of complementary data coming from different datasets into a unifying structure having a shared terminology and representation. Two initial OpenTox applications are presented as an illustration of the potential impact of OpenTox for high-quality and consistent structure-activity relationship modelling of REACH-relevant endpoints: ToxPredict which predicts and reports on toxicities for endpoints for an input chemical structure, and ToxCreate which builds and validates a predictive toxicity model based on an input toxicology dataset. Because of the extensible nature of the standardised Framework design, barriers of interoperability between applications and content are removed, as the user may combine data, models and validation from multiple sources in a dependable and time-effective way.

The cytochrome P(450) (CYP) system plays an integral part in the metabolism of drugs and other xenobiotics. Knowledge of the structural features required for interaction with any of the different isoforms of the CYP system is therefore immensely valuable in early drug discovery. In this paper, we focus on three major isoforms (CYP 1A2, CYP 2D6, and CYP 3A4) and present a data set of 335 structurally diverse drug compounds classified for their interaction (as substrate, inhibitor, or any interaction) with these isoforms. We also present machine learning models using a variety of commonly used methods (k-nearest neighbors, decision tree induction using the CHAID and CRT algorithms, random forests, artificial neural networks, and support vector machines using the radial basis function (RBF) and homogeneous polynomials as kernel functions). We discuss the physicochemical features relevant for each end point and compare it to similar studies. Many of these models perform exceptionally well, even with 10-fold cross-validation, yielding corrected classification rates of 81.7 to 91.9% for CYP 1A2, 89.2 to 92.9% for CYP 2D6, and 87.4 to 89.9% for CYP3A4. Our models help in understanding the structural requirements for CYP interactions and can serve as sensitive tools in virtual screenings and lead optimization for toxicological profiles in drug discovery.

We present a new approach to large-scale graph mining based on so-called backbone refinement classes. The method efficiently mines tree-shaped subgraph descriptors under minimum frequency and significance constraints, using classes of fragments to reduce feature set size and running times. The classes are defined in terms of fragments sharing a common backbone. The method is able to optimize structural inter-feature entropy as opposed to occurrences, which is characteristic for open or closed fragment mining. In the experiments, the proposed method reduces feature set sizes by >90 % and >30 % compared to complete tree mining and open tree mining, respectively. Evaluation using crossvalidation runs shows that their classification accuracy is similar to the complete set of trees but significantly better than that of open trees. Compared to open or closed fragment mining, a large part of the search space can be pruned due to an improved statistical constraint (dynamic upper bound adjustment), which is also confirmed in the experiments in lower running times compared to ordinary (static) upper bound pruning. Further analysis using large-scale datasets yields insight into important properties of the proposed descriptors, such as the dataset coverage and the class size represented by each descriptor. A final cross-validation run confirms that the novel descriptors render large training sets feasible which previously might have been intractable.

We propose a new kernel, based on 2-D structural chemical similarity, that integrates activity-specific information from the training data, and a new approach to applicability domain estimation that takes feature significances and activity distributions into consideration. The new kernel provides superior results than the well-established Tanimoto kernel, and activity-sensitive feature selection enhances prediction quality. Validation of local support vector regression models based on this kernel has been preformed with three publicly available datasets from the DSSTox project. One of them (Fathead Minnow Acute Toxicity) has been already modelled by other groups, and serves as a benchmark dataset, the other two (Maximum Recommended Therapeutic Dose, IRIS Lifetime Cancer Risk) have been modelled for the first time according to the knowledge of the authors. For all three models predictive accuracies increase with the prediction confidences that indicate the applicability domain. Depending on the confidence cutoff for acceptable predictions we were able to achieve90% predictions within 1 log unit of the experimental data for all datasets.

Different regulatory schemes worldwide, and in particular, the preparation for the new REACH (Registration, Evaluation and Authorization of CHemicals) legislation in Europe, increase the reliance on estimation methods for predicting potential chemical hazard. To meet the increased expectations, the availability of valid (Q)SARs becomes a critical issue, especially for endpoints that have complex mechanisms of action, are time-and cost-consuming, and require a large number of animals to test. Here, findings from the survey on (Q)SARs for mutagenicity and carcinogenicity, initiated by the European Chemicals Bureau (ECB) and carried out by the Istituto Superiore di Sanita' are summarized, key aspects are discussed, and a broader view towards future needs and perspectives is given.

The OECD has proposed five principles for validation of QSAR models used for regulatory purposes. Here we present a case study investigating how these principles can be applied to models based on Kohonen and counter propagation neural networks. The study is based on a counter propagation network model that has been built using toxicity data in fish fathead minnow for 541 compounds. The study demonstrates that most, if not all, of the OECD criteria may be met when modeling using this neural network approach.

This article describes the knowledge discovery process in predictive toxicology. This process consists of five major steps (i) feature calculation, (ii) feature selection, (iii) model induction, (iv) model validation and (v) interpretation of predictions and models. Data mining is a part of the knowledge discovery process and consists of the application of data analysis and discovery algorithms, which can be useful in all of the above steps. A brief review of suitable algorithms and their advantages and disadvantages is given for each knowledge discovery step, followed by a more detailed description of a problem-specific implementation of the lazar prediction system.

This paper explores the utility of data mining and machine learning algorithms for the induction of mutagenicity structure-activity relationships (SARs) from noncongeneric data sets. We compare (i) a newly developed algorithm (MOLFEA) for the generation of descriptors (molecular fragments) for noncongeneric compounds with traditional SAR approaches (molecular properties) and (ii) different machine learning algorithms for the induction of SARs from these descriptors. In addition we investigate the optimal parameter settings for these programs and give an exemplary interpretation of the derived models. The predictive accuracies of models using MOLFEA derived descriptors is similar to10- 15 %age points higher than those using molecular properties alone. Using both types of descriptors together does not improve the derived models. From the applied machine learning techniques the rule learner PART and support vector machines gave the best results, although the differences between the learning algorithms are only marginal. We were able to achieve predictive accuracies up to 78% for 10-fold cross-validation. The resulting models are relatively easy to interpret and usable for predictive as well as for explanatory purposes. ispartof: Journal of chemical information and computer sciences vol:44 issue:4 pages:1402-1411 ispartof: location:United States status: published