Data Mining in Drug Discovery: (57 Methods & Principles in Medicinal Chemistry)

Written for drug developers rather than computer scientists, this monograph adopts a systematic approach to mining scientifi c data sources, covering all key steps in rational drug discovery, from compound screening to lead compound selection and personalized medicine. Clearly divided into four sections, the first part discusses the different data sources available, both commercial and non-commercial, while the next section looks at the role and value of data mining in drug discovery. The third part compares the most common applications and strategies for polypharmacology, where data mining can substantially enhance the research effort. The final section of the book is devoted to systems biology approaches for compound testing. Throughout the book, industrial and academic drug discovery strategies are addressed, with contributors coming from both areas, enabling an informed decision on when and which data mining tools to use for one's own drug discovery project.

Table of Contents:
Preface A Personal Foreword PART ONE: Data Sources PROTEIN STRUCTURAL DATABASES IN DRUG DISCOVERY The Protein Data Bank: The Unique Public Archive of Protein Structures PDB-Related Databases for Exploring Ligand-Protein Recognition The sc-PDB, A Collection of Pharmacologically Relevant Protein-Ligand Complexes Conclusions PUBLIC DOMAIN DATABASES FOR MEDICINAL CHEMISTRY Introduction Databases of Small Molecule Binding and Bioactivity Trends in Medicinal Chemistry Data Directions Summary CHEMICAL ONTOLOGIES FOR STANDARDIZATION, KNOWLEDGE DISCOVERY, AND DATA MINING Introduction Background Chemical Ontologies Standardization Knowledge Discovery Data Mining Conclusions BUILDING A CORPORATE CHEMICAL DATABASE TOWARD SYSTEMS BIOLOGY Introduction Setting the Scene Dealing with Chemical Structures Increased Accuracy of the Registration of Data Implementation of the Platform Linking Chemical Information to Analytical Data Linking Chemicals to Bioactivity Data Conclusions PART TWO: Analysis and Enrichment DATA MINING OF PLANT METABOLIC PATHWAYS Introduction Pathway Representation Pathway Management Platforms Obtaining Pathway Information Constructing Organism-Specific Pathway Databases Conclusions THE ROLE OF DATA MINING IN THE IDENTIFICATION OF BIOACTIVE COMPOUNDS VIA HIGH-THROUGHPUT SCREENING Introduction to the HTS Process: The Role of Data Mining Relevant Data Architectures for the Analysis of HTS Data Analysis of HTS Data Identification of New Compounds via Compound Set Enrichment and Docking Conclusions THE VALUE OF INTERACTIVE VISUAL ANALYTICS IN DRUG DISCOVERY: AN OVERVIEW Creating Informative Visualizations Lead Discovery and Optimization Genomics USING CHEMOINFORMATICS TOOLS FROM R Introduction System Call Shared Library Call Wrapping Java Archives Conclusions PART THREE: Applications to Polypharmacology CONTENT DEVELOPMENT STRATEGIES FOR THE SUCCESSFUL IMPLEMENTATION OF DATA MINING TECHNOLOGIES Introduction Knowledge Challenges in Drug Discovery Case Studies Knowledge-Based Data Mining Technologies Future Trends and Outlook APPLICATIONS OF RULE-BASED METHODS TO DATA MINING OF POLYPHARMACOLOGY DATA SETS Introduction Materials and Methods Results Discussion Conclusion DATA MINING USING LIGAND PROFILING AND TARGET FISHING Introduction In Silico Ligand Profiling Methods Summary and Conclusions PART FOUR: System Biology Approaches DATA MINING OF LARGE-SCALE MOLECULAR AND ORGANISMAL TRAITS USING AN INTEGRATIVE AND MODULAR ANALYSIS APPROACH Rapid Technological Advances Revolutionize Quantitative Measurements in Biology and Medicine Genome-Wide Association Studies Reveal Quantitative Trait Loci Integration of Molecular and Organismal Phenotypes Is Required for Understanding Causative Links Reduction of Complexity of High-Dimensional Phenotypes in Terms of Modules Biclustering Algorithms Ping-Pong Algorithm Module Commonalities Provide Functional Insights Module Visualization Application of Modular Analysis Tools for Data Mining of Mammalian Data Sets Outlook SYSTEMS BIOLOGY APPROACHES FOR COMPOUND TESTING Introduction Step 1: Design Experiment for Data Production Step 2: Compute Systems Response Profiles Step 3: Identify Perturbed Biological Networks Step 4: Compute Network Perturbation Amplitudes Step 5: Compute the Biological Impact Factor Conclusions

About the Author :
Currently VP of Business Development at Prestwick Chemical SAS, Remy Hoffmann studied pharmacy at the University Louis Pasteur in Strasbourg, France, and gained his doctorate in medicinal chemistry. After 17 years spent at what is now Accelrys, where he worked on pharmacophore perception methods, he joined Thomson Reuters as a regional sales manager. Here he learnt the importance of curated scientific data, and the need to develop methods for mining this data so as to extract accurate information to support the decisionmaking process, and thus arrive at the knowledge stage. In his current role, Dr. Hoffmann oversees the deployment of Prestwick Chemical?s products and services to the drug discovery community, both in the pharma and biotech industries, as well as within the academic scientific community. Arnaud Gohier studied organic chemistry at the University of Le Mans and Nantes (France). He received his PhD in Molecular Modeling from the University of Joseph Fourier in Grenoble (France). In 1999, he joined the french pharmaceutical company Servier. Dr Gohier?s main areas of interest are drug design and chemoinformatics. Pavel Pospisil has been Manager of Computational Chemistry at Philip Morris International, R&D in Neuchatel, Switzerland, since 2008. He holds a BSc in biochemistry from the University of Joseph Fourier in Grenoble, France, and an MSc in biochemical engineering from the Institute of Chemical Technology in Prague, Czech Republic, and received his PhD in natural sciences from the Swiss Federal Institute of Technology (ETH), Zurich. He carried out his postdoctoral studies at ETH Zurich and with the pharmaceutical company, Arpida, now Evolva. In 2004, Dr. Pospisil became a postdoctoral fellow and research associate at Harvard Medical School, Boston, USA, where he focused on data mining for cancer targets and the discovery of low molecular radiolabeled cancer imaging analogs. In 2008, he took up a position as consultant at Hoffmann-La-Roche, Basel, Switzerland. His current interests are the automatic processing of molecules and computational toxicology.

Review :
“In summary, the book reflects the state-of-the-art for a rapidly changing field, with key emergent themes being the accessibility of public data, multiassay end-points for compounds, and the need to interpret these in the context of complex biological systems. It also usefully highlights some of the research challenges, with pointers to key likely future progress.”  (ChemMedChem, 1 June 2014)