Since the inception of this volume, the world's nancial climate has radically changed. Theemphasishasshiftedfromboomingeconomiesandeconomicgrowth totherealityofrecessionanddiminishingoutlook. Witheconomicdownturncomes opportunity,inallareasofchemistryfromresearchanddevelopmentthroughto productregistrationandriskassessment,replacementsarebeingsoughtforcostly time-consumingprocesses. Leadingamongstthereplacementsaremodelswithtrue predictivecapability. Ofthesecomputationalmodelsarepreferred. This volume addresses a broad need within various areas of the "chemical industries", from pharmaceuticals and pesticides to personal products to provide computationalmethodstopredicttheeffects,activitiesandpropertiesofmolecules. Itaddressestheuseofmodelstodesignnewmoleculesandassesstheirfateand effectsbothtotheenvironmentandtohumanhealth. Thereisanemphasisrunning throughoutthisvolumetoproducerobustmodelssuitableforpurpose. Thevolume aimstoallowthereaderto nddataanddescriptorsanddevelop,discoverandutilise validmodels. Gdansk, ' Poland TomaszPuzyn Jackson,MS,USA JerzyLeszczynski Liverpool,UK MarkT. D. Cronin May2009 CONTENTS Part I Theory of QSAR 1 QuantitativeStructure-ActivityRelationships(QSARs)- ApplicationsandMethodology...3 Mark T. D. Cronin 1. 1. Introduction...3 1. 2. PurposeofQSAR...4 1. 3. ApplicationsofQSAR...4 1. 4. Methods...5 1. 5. TheCornerstonesofSuccessfulPredictiveModels ...7 1. 6. AValidated(Q)SARoraValidPrediction? ...9 1. 7. UsinginSilicoTechniques ...9 1. 8. NewAreasforinSilicoModels...11 1. 9. Conclusions...11 References ...11 2 TheUseofQuantumMechanicsDerivedDescriptorsin ComputationalToxicology...13 Steven J. Enoch 2. 1. Introduction...13 2. 2. TheSchrodingerEquation...15 2. 3. Hartree-FockTheory...17 2. 4. Semi-EmpiricalMethods:AM1andRM1...18 2. 5. ABInitio:DensityFunctionalTheory...19 2. 6. QSARforNon-ReactiveMechanismsofAcute(Aquatic) Toxicity...19 2. 7. QSARsforReactiveToxicityMechanisms...21 2. 7. 1. AquaticToxicityandSkinSensitisation...21 2. 7. 2. QSARsforMutagenicity ...24 2. 8. FutureDirectionsandOutlook...25 2. 9. Conclusions...26 References ...26 vii viii Contents 3 MolecularDescriptors...29 Viviana Consonni and Roberto Todeschini 3. 1. Introduction...29 3. 1. 1. De nitions...29 3. 1. 2. History...31 3. 1. 3. Theoreticalvs. ExperimentalDescriptors...33 3. 2. MolecularRepresentation ...35 3. 3. TopologicalIndexes...38 3. 3. 1. MolecularGraphs...38 3. 3. 2. De nitionandCalculationofTopologicalIndexes(TIs) 39 3. 3. 3. Graph-TheoreticalMatrixes...42 3. 3. 4. ConnectivityIndexes ...48 3. 3. 5. CharacteristicPolynomial ...50 3. 3. 6. SpectralIndexes ...53 3. 4. AutocorrelationDescriptors ...

Since the inception of this volume, the world's nancial climate has radically changed. Theemphasishasshiftedfromboomingeconomiesandeconomicgrowth totherealityofrecessionanddiminishingoutlook. Witheconomicdownturncomes opportunity,inallareasofchemistryfromresearchanddevelopmentthroughto productregistrationandriskassessment,replacementsarebeingsoughtforcostly time-consumingprocesses. Leadingamongstthereplacementsaremodelswithtrue predictivecapability. Ofthesecomputationalmodelsarepreferred. This volume addresses a broad need within various areas of the "chemical industries", from pharmaceuticals and pesticides to personal products to provide computationalmethodstopredicttheeffects,activitiesandpropertiesofmolecules. Itaddressestheuseofmodelstodesignnewmoleculesandassesstheirfateand effectsbothtotheenvironmentandtohumanhealth. Thereisanemphasisrunning throughoutthisvolumetoproducerobustmodelssuitableforpurpose. Thevolume aimstoallowthereaderto nddataanddescriptorsanddevelop,discoverandutilise validmodels. Gdansk, ' Poland TomaszPuzyn Jackson,MS,USA JerzyLeszczynski Liverpool,UK MarkT. D. Cronin May2009 CONTENTS Part I Theory of QSAR 1 QuantitativeStructure-ActivityRelationships(QSARs)- ApplicationsandMethodology...3 Mark T. D. Cronin 1. 1. Introduction...3 1. 2. PurposeofQSAR...4 1. 3. ApplicationsofQSAR...4 1. 4. Methods...5 1. 5. TheCornerstonesofSuccessfulPredictiveModels ...7 1. 6. AValidated(Q)SARoraValidPrediction? ...9 1. 7. UsinginSilicoTechniques ...9 1. 8. NewAreasforinSilicoModels...11 1. 9. Conclusions...11 References ...11 2 TheUseofQuantumMechanicsDerivedDescriptorsin ComputationalToxicology...13 Steven J. Enoch 2. 1. Introduction...13 2. 2. TheSchrodingerEquation...15 2. 3. Hartree-FockTheory...17 2. 4. Semi-EmpiricalMethods:AM1andRM1...18 2. 5. ABInitio:DensityFunctionalTheory...19 2. 6. QSARforNon-ReactiveMechanismsofAcute(Aquatic) Toxicity...19 2. 7. QSARsforReactiveToxicityMechanisms...21 2. 7. 1. AquaticToxicityandSkinSensitisation...21 2. 7. 2. QSARsforMutagenicity ...24 2. 8. FutureDirectionsandOutlook...25 2. 9. Conclusions...26 References ...26 vii viii Contents 3 MolecularDescriptors...29 Viviana Consonni and Roberto Todeschini 3. 1. Introduction...29 3. 1. 1. De nitions...29 3. 1. 2. History...31 3. 1. 3. Theoreticalvs. ExperimentalDescriptors...33 3. 2. MolecularRepresentation ...35 3. 3. TopologicalIndexes...38 3. 3. 1. MolecularGraphs...38 3. 3. 2. De nitionandCalculationofTopologicalIndexes(TIs) 39 3. 3. 3. Graph-TheoreticalMatrixes...42 3. 3. 4. ConnectivityIndexes ...48 3. 3. 5. CharacteristicPolynomial ...50 3. 3. 6. SpectralIndexes ...53 3. 4. AutocorrelationDescriptors ...

The development of computational methods that support human health and environmental risk assessment of engineered nanomaterials (ENMs) has attracted great interest because the application of these methods enables us to fill existing experimental data gaps. However, considering the high degree of complexity and multifunctionality of ENMs, computational methods originally developed for regular chemicals cannot always be applied explicitly in nanotoxicology. This book discusses the current state of the art and future needs in the development of computational modeling techniques for nanotoxicology. It focuses on (i) computational chemistry (quantum mechanics, semi-empirical methods, density functional theory, molecular mechanics, molecular dynamics), (ii) nanochemoinformatic methods (quantitative structure-activity relationship modeling, grouping, read-across), and (iii) nanobioinformatic methods (genomics, transcriptomics, proteomics, metabolomics). It reviews methods of calculating molecular descriptors sufficient to characterize the structure of nanoparticles, specifies recent trends in the validation of computational methods, and discusses ways to cope with the uncertainty of predictions. In addition, it highlights the status quo and further challenges in the application of computational methods in regulation (e.g., REACH, OECD) and in industry for product development and optimization and the future directions for increasing acceptance of computational modeling for nanotoxicology.