Название: Big Data Analytics in Chemoinformatics and Bioinformatics: With Applications to Computer-Aided Drug Design
Автор: Subhash C. Basak, Marjan Vracko
Издательство: Elsevier
Год: 2023
Страниц: 434
Язык: английский
Формат: pdf
Размер: 20.0 MB

Big Data Analytics in Chemoinformatics and Bioinformatics: With Applications to Computer-Aided Drug Design, Cancer Biology, Emerging Pathogens and Computational Toxicology provides an up-to-date presentation of big data analytics methods and their applications in diverse fields. The proper management of Big Data for decision-making in scientific and social issues is of paramount importance. This book gives researchers the tools they need to solve Big Data problems in these fields. It begins with a section on general topics that all readers will find useful and continues with specific sections covering a range of interdisciplinary applications.

Here, an international team of leading experts review their respective fields and present their latest research findings, with case studies used throughout to analyze and present key information.

The first section, General Section, of the book has three chapters. Chapter 1 briefly traces the history of the development of chemodescriptors and biodescriptors spanning three centuries—from the eighteenth century to the present. Chapter 2 deals with the problem of robust model building from noisy high-dimensional data, focusing primarily on the robustness aspects against data contamination. The author also demonstrates the utility of his method in the prediction of salmonella mutagenicity of a set of amines, a class priority pollutants. Chapter 3 delves into the ethical issues associated with the landscape of desirable qualities such as fairness, transparency, privacy, and robustness of currently used Machine Learning (ML) methods of Big Data analysis.

The second section, Chemistry and Chemoinformatics Section, of the book has nine chapters. Chapter 4 discusses the use of Big Data in the characterization of adverse outcome pathways (AOPs), a novel paradigm in toxicology. The author integrated “big data”-the -omics and high-throughput (HT) screening data—to derive AOPs for chemical carcinogens. Chapter 5 discusses the latest progress in the use of ML and DL (Deep Learning) methods in creating systems that automatically mine patterns and learn from data. The author also discuss the challenges and usefulness of DL for quantitative structure-activity relationship (QSAR) modeling. Chapter 6 describes retrosynthetic planning and analysis of organic compounds in the synthetic space using Big Data sets and in silico algorithms. Chapter 7 discusses that the vast amount of historical chemical information is not only a rich source of data, but also a useful tool for studying the evolution of chemistry, chemoinformatics, and bioinformatics through a computational approach to the history of chemistry. The author exemplifies that by a case study of recent results on the computational analysis of the evolution of the chemical space. Chapter 8 gives a detailed description of combinatorial techniques useful in studying large data sets with hypercubes and halocarbons as the main focus. Quantum chemical techniques discussed here can generate electronic parameters that have potential for use in QSAR for toxicity prediction of Big Data sets. Chapter 9 deals with the use of computed high-level quantum chemical descriptors derived from the density functional theory in the prediction of property/toxicity of chemicals. Chapter 10 covers the important area of the use of computed pharmacophores in practical drug design from analysis of large databases. Chapter 11 uses ML based classification methods for the detection of hot spots in protein-protein interactions and prediction of new hotspots. Chapter 12 discusses applications of decision tree methods like recursive partitioning, phylogenetic-like trees, multidomain classification, and fuzzy clustering within the context of small molecule drug discovery from analysis of large databases.

The third section, Bioinformatics and Computatioanl Toxicology Section, of the book has seven chapters. Chapter 13 discusses their contributions in the emerging area of mathematical proteomics approach in developing biodescriptors for the characterization of bioactivity and toxicity of drugs and pollutants. Chapter 14 discusses the important role of efficient computational frameworks developed to catalog and navigate the protein space to help the drug discovery process. Chapter 15 discusses applications of ML and DL approaches to HT sequencing data in the development of precision medicine using single-nucleotide polymorphisms as a tool of reference. Chapter 16 discusses the development and use of a new class of sequence comparison methods based on alignment-free sequence descriptors in the characterization of emerging global pathogens. Chapter 17 discusses the important and emerging issue of different ways of building QSARs from large and diverse data sets that can be continuously updated and expanded over time. The importance of modularity in scalable QSAR system development is also discussed. Chapter 18 deals with the applications of network analysis and Big Data to study interactions of drugs with their targets in the biological systems. The authors point out that a paradigm shift integrating big data and complex network is needed to understand the expanding universe of drug molecules, targets, and their interactions. Finally, Chapter 19 reports the use of ML approaches consisting of supervised and unsupervised techniques in the analysis of RNA sequence data of breast cancer to derive important biological insights.

Finally, we would like to specially mention that in drug research and toxicology, we are witnessing an explosion of data, which are expressed by four principal Vs - volume, velocity, variety, and veracity. However, the data per se is useless, the real challenge is the transition to the last two steps on the three-step path to knowledge: data - information - knowledge. A future challenge for us is to integrate both data platforms—big and small—into a new and integrated knowledge extraction system.

- Brings together the current knowledge on the most important aspects of Big Data, including analysis using deep learning and fuzzy logic, transparency and data protection, disparate data analytics, and scalability of the Big Data domain
- Covers many applications of Big Data analysis in diverse fields such as chemistry, chemoinformatics, bioinformatics, computer-assisted drug/vaccine design, characterization of emerging pathogens, and environmental protection
- Highlights the considerable benefits offered by Big Data analytics to science, in biomedical fields and in industry