Computational Toxicology: Methods and Protocols (Methods in Molecular Biology, 2834)

Dedication
Preface
Contents
Contributors
Part I: The Tools of the Game
Chapter 1: QSAR: Using the Past to Study the Present
	1 Introduction
	2 Building QSAR Models
		2.1 Biology
		2.2 Chemistry
		2.3 Modeling
	3 Evolution and Trends in QSAR Modeling
	4 QSAR as Induction: From Data to Knowledge
		4.1 Mathematical Methods: Statistics, Probability, and Machine Learning
		4.2 Knowledge, Models, and Theory
	5 Epistemology of QSAR
		5.1 Validation
		5.2 Justification
		5.3 Interpretability
	6 QSAR for Endocrine Disruptors: A Case Study
	7 Final Considerations and Next Challenges
	References
Chapter 2: Molecular Similarity in Predictive Toxicology with a Focus on the q-RASAR Technique
	1 Introduction
		1.1 History of the Development of Various Similarity and Distance Measures
		1.2 Local vs Global Similarity
		1.3 Human Perspective and Consideration
	2 QSAR and Read-Across: An Overview
		2.1 Some of the Available Tools for the Read-Across-Based Predictions
	3 Combining QSAR and Read-Across: Development of the Novel q-RASAR Modeling Approach
		3.1 Methodology for the Calculation of the RASAR Descriptors
		3.2 The Concept of Leave-Same-Out (LSO) Algorithm to Reduce Overfitting of the Calculated RASAR Descriptors
		3.3 Data Fusion and Model Development
		3.4 The List of RASAR Descriptors Proposed by Roy´s Group
		3.5 Determination of the Applicability Domain of the RASAR Models: Development of the DTC Plot to Identify the Prediction Conf...
		3.6 Applications of q-RASAR
	4 Conclusion
	References
Chapter 3: Weight of Evidence: Criteria and Applications
	1 Introduction
	2 The Used Non-testing Methods
	3 Case Study
		3.1 Bioaccumulation
		3.2 Conclusion on the First Example
		3.3 Case Study with SWAN to Assess the Carcinogenicity
	References
Chapter 4: Integration of QSAR and NAM in the Read-Across Process for an Effective and Relevant Toxicological Assessment
	1 Introduction
	2 Problem Formulation
	3 Characterization of the Target Substance
	4 QSAR: Initial RAx Hypothesis
	5 Source Compound Identification
	6 RAx Evaluation
	7 Generation of NAMs Data
	8 Applicability Domain
	9 Evaluation of Uncertainties
	10 Conclusion
	References
Part II: Molecular and Data Modeling
	Chapter 5: Automated Workflows for Data Curation and Machine Learning to Develop Quantitative Structure-Activity Relationships
		1 Introduction
		2 Materials
			2.1 Data Curation Workflow
			2.2 QSAR Development Workflow
		3 Methods
			3.1 Set Up the Workflows
				3.1.1 Import the Workflow
				3.1.2 Install  cddd
				3.1.3 Install VSURF
			3.2 Execute the Data Curation Workflow
				3.2.1 Part 1
				3.2.2 Manual Check
				3.2.3 Part 2
				3.2.4 Execute the Workflow-SMILES Input
			3.3 Execute the QSAR Workflow
				3.3.1 QSAR Development
				3.3.2 Prediction of External Chemicals
		4 Notes
		References
	Chapter 6: Applicability Domain for Trustable Predictions
		1 Introduction
		2 Understanding the Applicability Domain
			2.1 Definition
			2.2 Methods
		3 Applicability Domain in Machine Learning and Artificial Intelligence
			3.1 Defining AD in AI and ML
			3.2 Theoretical Basis of Applicability Domain in AI and ML
			3.3 Measurement of Applicability Domain in AI and ML
				3.3.1 DA Index (κ, γ, δ)
				3.3.2 Class Probability Estimation
				3.3.3 Class Probability Estimates Using the Local Vicinity
				3.3.4 Confidence Measure and Class Probability Estimates from Boosting
				3.3.5 Class Probability Estimates from Classification Neural Networks
				3.3.6 PROB-STD
				3.3.7 Cosine (cosα)
				3.3.8 Subgroup Discovery (SGD)
			3.4 AD in QSAR
			3.5 AD in RASAR and Q-RASAR
		4 Applications of the AD in ML and AI
		5 Overview
		References
	Chapter 7: Approaching Pharmacological Space: Events and Components
		1 Introduction
		2 Methods
		3 Applications of the Pharmacological Space in Virtual Screening Campaigns
			3.1 Setting the Scene: The Muscarinic Receptors
			3.2 The Fertile Case of the TRPM8 Channel
			3.3 The Isomeric Space as Applied to the M-Pro Enzyme of Sars-CoV-2
		4 Applications of the Pharmacological Space in Toxicity Predictions: The hERG-Related Cardiotoxicity
		5 Conclusions
		References
	Chapter 8: The Potential of Molecular Docking for Predictive Toxicology
		1 Introduction
		2 Materials
			2.1 3D Protein Structures
			2.2 Ligand Structures
			2.3 Structures Preparation
			2.4 Docking Software
			2.5 Scoring Functions (SF)
		3 Methods
			3.1 Protein and Ligand Selection
			3.2 Structures Preparation
			3.3 Binding Site Definition
			3.4 Docking with  Gold
			3.5 Docking with Autodock (See Note 2)
			3.6 Rescoring with  HINT
			3.7 Consensus Scoring and 3D Analysis
		4 Notes
		References
	Chapter 9: Computational Toxicology Methods in Chemical Library Design and High-Throughput Screening Hit Validation
		1 Introduction
		2 Applications
			2.1 Screening Library Development Tools
			2.2 Tools for Prediction of Reactivity/Toxicity
			2.3 Structure Alerts for Reactive and/or Toxic Functional Groups
			2.4 Artificial Intelligence and Machine Learning Algorithms
		3 Summary of Reactive Structure Filters
		4 Conclusions
		References
Part III: Applicative Examples of Predictive Toxicology
	Chapter 10: Toxicity Potential of Nutraceuticals
		1 Introduction
		2 Common Nutraceuticals
		3 Nutraceuticals with a Toxic Potential
			3.1 Cannabis/Hemp
			3.2 Goldenseal
			3.3 Ginkgo biloba
			3.4 Green Tea Extract
			3.5 Green Coffee Bean/Caffeine
			3.6 Garcinia cambogia
			3.7 Kava
			3.8 St. John´s  Wort
			3.9 Bitter Melon
			3.10 Aloe  vera
			3.11 Ephedrine Alkaloids
			3.12 Pennyroyal  Oil
		4 Nutraceuticals´ Safety Concern During Perioperative  Care
		5 Toxic Contaminants in Nutraceuticals
			5.1 Pyrrolizidine Alkaloids
			5.2 Mycotoxins
			5.3 Heavy Metals
			5.4 Pesticides
		6 Models for Nutraceuticals´ Efficacy, Safety, and Toxicity Assessment
		7 Nutraceutical-Drug Interaction and Toxicity Outcome
		8 Biomarkers of Nutraceuticals´ Toxic Potential
		9 Management of Nutraceuticals´ Toxicity
		10 Concluding Remarks and Future Directions
		References
	Chapter 11: Development, Use, and Validation of (Q)SARs for Predicting Genotoxicity and Carcinogenicity: Experiences from Ital...
		1 Introduction
		2 Methods
			2.1 (Q)SAR Development
				2.1.1 Database Development
			2.2 (Q)SAR  Use
				2.2.1 (Q)SAR Use in a Regulatory Setting: Derivation of Toxicity Thresholds
				2.2.2 (Q)SAR Use in Applied Research: Study the Impact of Structural Changes on Chemical Hazard
			2.3 (Q)SAR Evaluation
				2.3.1 The (Q)SAR Assessment Framework
		References
	Chapter 12: Adverse Outcome Pathways Mechanistically Describing Hepatotoxicity
		1 Introduction
		2 Structure
		3 Development and Assessment
			3.1 Background
			3.2 Development
			3.3 Assessment
			3.4 Quantification
		4 AOPs on Liver Toxicity
			4.1 Cholestasis
			4.2 Liver Steatosis
			4.3 Liver Fibrosis
			4.4 Liver Cancer
		5 Applications
			5.1 Development of Quantitative Structure-Activity Relationships
			5.2 Grouping of Chemicals into Chemical Categories
			5.3 Elaboration of Prioritization Strategies
			5.4 Development of Testing Strategies
		6 Conclusions and Perspectives
		References
	Chapter 13: Machine Learning in Early Prediction of Metabolism of Drugs
		1 Introduction
		2 The Importance to Predict Metabolism in Toxicology
		3 Machine Learning
			3.1 Deep Learning
		4 The Prediction of Metabolism
		5 From Molecular Structures to Sites of Metabolites
		6 Graph-Based Methods to Train Models
		7 Several Metabolisms to Predict: Some Practical Perspectives
		8 Reliability and Regulatory Acceptance
		9 Discussion and Conclusions
		References
	Chapter 14: In Vitro Cell-Based MTT and Crystal Violet Assays for Drug Toxicity Screening
		1 Introduction
		2 Materials
			2.1 Equipment
			2.2 MTT Assay
			2.3 Reagents for Crystal Violet Assay
		3 Methods
			3.1 Experimental Design
			3.2 Drug Solution
			3.3 Cell Culture and Drug Treatment
			3.4 Protocol 1: MTT Assay (see Fig. 3)
			3.5 Protocol 2: Crystal Violet Assay (see Fig. 3)
			3.6 Data Analysis
		4 Notes
		References
	Chapter 15: Recent Advances in Nanodrug Delivery Systems Production, Efficacy, Safety, and Toxicity
		1 Introduction
		2 Materials and Methods of Production of  NPs
			2.1 The Innovative Microfluidic Technique Applied to NP Production
			2.2 Treatment, Manufacture Impurities, and Stability of  NPs
		3 Targets and Mechanism of Interactions with Cellular and Intracellular Systems of  NPs
		4 In Vitro, In Vivo, and In Silico Nanotoxicological Assays
			4.1 In Vitro and In Vivo Case Studies for Specific  NPs
		5 How Have the COVID-19 Vaccines Been Developed So Quickly? Does This Mean that Their Safety, Efficacy, and Quality Have Been ...
		6 Final Considerations
		References
	Chapter 16: Investigating the Benefit-Risk Profile of Drugs: From Spontaneous Reporting Systems to Real-World Data for Pharmac...
		1 Introduction
			1.1 Aim and Methods of Pharmacovigilance
		2 Passive Surveillance
			2.1 Spontaneous Reporting Databases
			2.2 Analysis of Spontaneous Reporting of Suspected ADRs
			2.3 Case Studies
				2.3.1 Evaluation of CAR-T Cell Therapy and Kidney Failure from the WHO Safety Database Vigibase
				2.3.2 Evaluation of Immune Checkpoint Inhibitors and Cardiac ADRs from the European Database Eudravigilance
				2.3.3 Evaluation of Direct Oral Anticoagulant Safety Profile from the Italian National Pharmacovigilance Network
				2.3.4 Evaluation of Antiseizure Medications and Pediatric ADRs from the Italian National Pharmacovigilance Network
		3 The Role of Real-World Data and Real-World Evidence in Pharmacovigilance
			3.1 Healthcare Database and Record Linkage
			3.2 Case Studies
				3.2.1 Post-Marketing Surveillance of Biological Drugs from an Italian Multidatabase Network
			3.3 Social Networks
		4 Conclusions
		References
	Chapter 17: MolPredictX: A Pioneer Mobile App Version for Online Biological Activity Predictions by Machine Learning Models
		1 Introduction
		2 Web Tools Using Machine Learning for Chemical Compound Prediction
			2.1 MolpredictX
			2.2 Chembench
			2.3 MuDRA
			2.4 PASS
			2.5 SwissADME
			2.6 SwissTargetPrediction
			2.7 OCHEM
			2.8 MuSSel
			2.9 DeepChem
		3 Mobile Apps for Chemical Compound Prediction
			3.1 MolpredictX
			3.2 ChemDoodle
			3.3 APP
			3.4 Open Chemistry
			3.5 Open-Source Bayesian Models
			3.6 KingDraw: Free Chemical Structure Editor
			3.7 Reaction Flash
			3.8 Chemical Equation Balancer
			3.9 AppsPred
		4 MolpredictX
		5 MolpredictX: App Mobile
		6 MolPredictX: App Utility
		7 Using MolpredictX App for Predicting Biological Activities: Test Cases
		8 Conclusion and Perspectives
		References
	Chapter 18: TIRESIA and TISBE: Explainable Artificial Intelligence Based Web Platforms for the Transparent Assessment of the D...
		1 Introduction
		2 TIRESIA and TISBE: Web Based Platforms for Predicting Dev  Tox
		3 TIRESIA and TISBE: Real-Life Applications
		4 Chemical and Feature Representation
		5 The Learning Framework
		6 Explainability Analysis
		7 Applicability Domain
		8 Case Study 1: Estradiol Profiling
		9 Case Study 2: Aztreonam Profiling
		10 Conclusions
		References
	Chapter 19: PFAS-Biomolecule Interactions: Case Study Using Asclepios Nodes and Automated Workflows in KNIME for Drug Discover...
		1 Introduction
		2 The Asclepios KNIME Nodes for In Silico Investigations
		3 The Asclepios KNIME Workflow: The PPARα-PFAS Example
			3.1 Ligand Preparation
			3.2 Receptor Preparation
			3.3 Molecular Docking Calculations
			3.4 Molecular Dynamics Simulations
		4 Binding Energy Calculations
		5 Future Directions
		6 Conclusions
		References
Index