Artificial Intelligence in Bioinformatics and Chemoinformatics

✍ Scribed by Yashwant Pathak (editor), Surovi Saikia (editor), Sarvadaman Pathak (editor), Jayvadankumar Patel (editor)

• Publisher: CRC Press
• Year: 2023
• Tongue: English
• Leaves: 275
• Edition: 1
• Category: Library

✦ Synopsis

The authors aim to shed light on the practicality of using machine learning in finding complex chemoinformatics and bioinformatics applications as well as identifiying AI in biological and chemical data points. The chapters are designed in such a way that they highlight the important role of AI in chemistry and bioinformatics particularly for the classification of diseases, selection of features and compounds, dimensionality reduction and more. In addition, they assist in the organization and optimal use of data points generated from experiments performed using AI techniques. This volume discusses the development of automated tools and techniques to aid in research plans.

Features

• Covers AI applications in bioinformatics and chemoinformatics

• Demystifies the involvement of AI in generating biological and chemical data

• Provides an Introduction to basic and advanced chemoinformatics computational tools

• Presents a chemical biology based toolset for artificial intelligence usage in drug design
  
• Discusses computational methods in cancer, genome mapping, and stem cell research

✦ Table of Contents

Cover
Half Title
Title Page
Copyright Page
Table of Contents
Preface
Editors
List of Contributors
1 Bridging Bioinformatics and Chemoinformatics Approaches in Public Databases
1 Artificial Intelligence and Bioinformatics
1.1 Introduction to Bioinformatics and Its Applications
1.2 Artificial Intelligence for Biology
1.3 Applications of AI in Solving Biological Problems
1.3.1 The Protein-Folding Problem
1.3.2 AI in Disease Biology and Personalized Medicine
1.3.3 The Problem of Identifying Genomic Variants From Sequence Reads
1.3.4 Predicting the Transcription Factor-Binding Sites
2 Artificial Intelligence and Chemoinformatics
2.1 Introduction to Chemoinformatics
2.1.1 Drug Discovery
2.1.2 Computer-Assisted Synthesis Design
2.2 Chemoinformatics: From Big Data to Artificial Intelligence
2.2.1 Artificial Intelligence
2.2.2 Deep Neural Network
2.3 Application of Chemoinformatics and AI in Drug Discovery
3 Bridging Bioinformatics and Chemoinformatics Approaches
3.1 Bioinformatics Approaches
3.1.1 Human Health
3.1.2 Evolutionary Biology
3.2 Chemoinformatics Approaches
3.3 Bioinformatics and Chemoinformatics at an Interface With Systems Biology
4 Publicly Available Resources in Bioinformatics and Cheminformatics
4.1 Data and Knowledge Management
4.2 Data Extraction and Transformation
4.3 Big Data Analysis
5 Biochemoinformatics: Integrating Bioinformatics and Chemoinformatics
5.1 Genomics
5.2 Proteomics
5.3 Complex Chemical Structures
6 Artificial Intelligence in Healthcare
6.1 Implications of AI ML
6.2 AI in Bioinformatics
6.3 Transforming Healthcare With AI
6.4 AI in Drug Design and Development
6.5 Future Prospects and Challenges
7 Conclusion
References
2 An Introduction to Basic and Advanced Chemoinformatics Computational Tools
1 Introduction
2 Cheminformatics Application in Drug Discovery
2.1 Selectivity and ADMET Properties
2.2 Optimization of Compounds
2.3 Predictivity Against Structural Diversity
2.4 Data Mining
3 Cheminformatics in Database Creation
3.1 Molecular Descriptors
3.2 Similarity Index
3.2.1 Construction of Chemical Space
3.3 Specificity
4 Advances in Chemoinformatics Computational Tools
4.1 SMILES
4.2 DeepSMILES
4.3 IncHI
4.4 SELFIES
5 Addition to PubChem Information
5.1 Expandability in Search Terms
5.1.1 Integration With Entrez for Textual Search
5.1.2 PubChem Search Using Non-Textual Query
6.1.1 PubChem3D
6.1.2 ChemSpider
6.1.3 3D-E-Chem-VM
6.1.4 LigandBox
6 Database of 3D Structures
6.2 Dealing With Randomness of Compounds
6.3 RSS Feeds
7 Chemoinformatics With Semantic Web Technologies
7.1 Integration and Management of Data
7.2 Metadata
7.3 Lexis in Chemoinformatics
7.4 Access and Discovery Levels
7.4 Linking Data
8 Future Prospects in Chemoinformatics
9 Conclusions
Conflict of Interest
Acknowledgments
References
3 An Introduction to Basic and Advanced Bioinformatics Computational Tools
1 Introduction
2 Sequence Analysis Tools
2.1 Sequence Alignment Or Reference-Based Mapping/Visualization Tools
2.2 Alignment-Free Sequence Analysis Tools
2.3 De Novo Sequence Assembly Tools
2.4 Gene Prediction Tools
2.5 Nucleic Acid Design and Simulation Tools
2.6 DNA Melting (Denaturation) Prediction Tools
2.7 Phylogenetics Tools
3 Structure Prediction Tools
3.1 Protein Structure Prediction Tools
3.2 Nucleic Acid Structure Prediction Tools
4 Omics Tools
4.1 Genomics
4.2 Transcriptomics
4.3 Proteomics
4.4 Metabolomics
4.5 Other Omics
4.6 Pathway Enrichment Analysis
5 Challenges and Perspective
References
4 Computational Methods in Cancer, Genome Mapping, and Stem Cell Research
1 Introduction
2 Genome Mapping as a Tool for Cancer Research
3 Stem Cells and Cancer Research
3.1 Cancer Stem Cell Research and Heterogeneity of Tumor
3.2 Anticancer Vaccines Based On Stem Cells
4 Recent Advances in Stem Cell Research and Genome Mapping for Cancer
5 Limitations of Stem Cell Research and Genome Mapping for Cancer Research
6 Conclusion
References
5 Using Chemistry to Understand Biology
Abbreviations
1 INTRODUCTION
1.1 Origin of Science
1.2 Studying Biology
1.3 Artificial Intelligence
2 Understanding Biology in Purview of Chemistry
2.1 Atoms and Molecules
2.2 Biomolecules
2.2.1 Protein Architecture
2.3 Water
2.3.1 Polarity
2.3.2 Hydrogen Bonding
2.3.3 Cohesion
2.3.4 Surface Tension
3 The Chemistry of Biology
3.1 Knowing Living Molecules W.s.r. Chemistry
3.2 Biochemical Processes
3.2.1 Formation of Lactic and Muscle Fatigue
3.2.2 Breakdown of Glucose in the Presence of Oxygen
3.2.3 The Fate of Pyruvate
3.3 Plant Kingdom
3.3.1 Photosynthesis
3.3.2 Cellular Respiration
4 Drug Discovery and Drug–Receptor Interaction
4.1 How Does This Interaction Occur?
4.2 Typically, Computer-Aided Drug Design Consists of the Following Process
4.3 Apart From This, Machine Learning-Based Procedure Is Also Popular. It Consists of the Following Steps.
5 Conclusion
References
6 Chemical Biology-Based Toolset for Artificial Intelligence Usage in Drug Design
1 Introduction
1.1 Chemical Biology
1.2 Drug Design
1.3 Artificial Intelligence
2 Artificial Intelligence in Drug Design
3 Chemical Biology Toolset for AI in Drug Design
3.1 Chemical Probes
3.2 Phenotypic Screening
3.3 Antisense Oligonucleotide (ASO) and RNAis
References
7 Machine Learning and Data Mining: Uses and Challenges in Bioinformatics
1 Introduction
2.1 Machine Learning
2.2 Review of Research Using Machine Learning
a) Clinical Research
b) Ecosystem Service Research
c) Healthcare
d) Traffic Prediction
e) Agriculture
f) ML in Personalised Medicines
g) Drug Development
h) Acute Illness Detection
2.3 ML Techniques Employed in Bioinformatics (Figure 7.6.)
a. Dimensionality Reduction
b. Clustering
c) Deep Learning
d) Classification/Regression
3.1 Data Mining
3.2 Differentiation Between Data Mining and Statistics
3.3 Data Mining Techniques
3.4 Classification of Data Mining
a) Regression
b) Time Series Analysis
c) Clustering
d) Sequence Discovery Analysis
e) Summarization Analysis
f) Association Rules Analysis
g) Predictive Data Mining
4 Role of Machine Learning and Data Mining in Bioinformatics
5 Challenges in the Handling of Big Data
5.1 Cleansing, Obtaining, and Capturing Data
5.2 Storage, Distribution, and Transfer
5.3 Analysis and Result Gathering
5.4 Ethics-Related Matters
6 Conclusion
References
8 Application of Deep Learning in Chemistry and Biology
LIST OF ABBREVIATIONS
1 Introduction to Deep Learning
1.1 History of Deep Learning
1.2 How and When to Use Deep Learning
1.3 The Position of Deep Learning in AI
2 Deep Learning and Its Application
2.1 Application of DL to Predict Compound Properties and Activities
2.2 De Novo Design Using DL
2.3 Use of DL for Retrosynthetic Assessment and Response Prediction
2.4 Application of Convolutional Neural Networks to Forecast Ligand–Protein Interactivities
2.5 Investigation of Bioimaging Using Deep Learning
2.6 DL for Drug Development
3 Application of Deep Learning in Chemistry
3.1 Accelerated Computational Models
3.2 Drug Design
3.3 Designing Molecules
3.4 Applications in Atomistic Representations
3.5 Reaction Outcome Prediction
3.6 Molecule Property Prediction
4 Tools and Methods Used in Chemistry
4.1 DTI-CNN
4.2 DeepCPI
4.3 WideDTA
4.4 PADME
4.5 DeepAffinity
4.6 DeepBAR
4.7 Methodology
Expert Systems in Medicine
Key Features of an ES Are
FL
AI-ONE
AI in Designing Drug Molecule
4.8 Deep Learning Algorithms
Types of Algorithms Used in Deep Learning
5 Basics of Deep Learning in Biology
5.1 Application of Deep Learning in Biology With Examples
• Omics
• Genomics
• Proteomics
• Metabolomics
• Bioimaging
• Image Restoration
• Image Partitioning
• Image Quantification
• Deep Learning for Segmentation
• Deep Learning in (Body/Brain) Interfaces
• Deep Learning in the Diagnosis of Disease
6 Tools, Methods, and Algorithm Used In Biology
• RNA-Protein–binding Sites Prediction With CNN
• CNN and RNN for DNA Sequence Function Estimation
• Transfer Learning and ResNet By Using Biomedical Image Classification
• Graph Embedding for Novel Protein Interaction Prediction Using GCN
• Biology Image Super-Resolution Using GAN
• VAE Used for High-Dimensional Biological Data Embedding and Generation
7 Goals and Advancements
8 Challenges and Outlook
8.1 Technical Challenges
8.2 Data Security and Storage
8.3 Legal Issues
8.4 Economical, Political, Social, Ethical, and Legal Aspects
9 Conclusion
References
9 Text Mining in Chemistry for Organizing Chemistry Data
1 Introduction
2 What Are Chemical Information Researchers Looking For?
3 Where Do Researchers Find Chemical Information?
3.1 Patents
3.2 Scholarly Articles
3.3 Regulatory Reports
4 Text Mining of Chemistry Literature
4.1 Text Collection
4.2 Document Processing
4.3 Text Representation
4.3.1 Sentence Splitting
4.3.2 Tokenization
4.3.3 Part-Of-Speech Tagging
4.3.4 Chemical Named Entry Recognition
4.3.5 Phase Parsing
4.4 Information Extraction
5 Cheminformatics
6 Application
7 Challenges
8 Future Trends and Perspectives
9 Conclusion
References
10 CNN Use and Performance in Virtual Screening
1 Introduction
2 Virtual Screening
2.1 Structure-Based Methods
2.2 Ligand-Based Methods
3 Artificial Intelligence-Based Virtual Screening
3.1 Molecular Formats
3.2 Chemical Representation
4 Artificial Neural Networks
5 Challenges Addressed By Neural Networks
6 Convolutional Neural Networks
6.1 Components of a CNN
6.1.1 Convolutional Layers
6.1.2 Fully Connected Multilayer Perceptron
6.1.3 Pooling Layers
7 Applications of CNN in Virtual Screening
7.1 Prediction of Binding Affinity
7.2 Protein–Ligand-Binding CNN Models
7.3 Drug–Target-Binding Prediction
8 Challenges and Future Outlook
9 Conclusion
References
11 Machine Learning in Improving Force Fields of Molecular Dynamics
Molecular Dynamics
What Are These Forces?
Parameterization
Molecular Simulation Problems That Can Be Addressed By Machine Learning
1 Potential Energy Surfaces
2 Free Energy Surfaces
3 Coarse Graining
4 Kinetics
5 Sampling and Thermodynamics
Inclusion of Physics Into Machine Learning for Molecular Dynamics and Simulations
1 Data Augmentation
2 Building Physical Constraints Into Machine Learning Model
3 Invariance and Equivariance
4 Parameter Sharing and Convolutions
Machine Learning Frameworks for Molecular Simulations
1 Deep Potential Net, Behler-Parrinello, and ANI
2 Deep Tensor Neural Network, SchNet, and Continuous Convolutions
3 Coarse Graining CGnets
4 Kinetics VAMPnets
5 Sampling/Thermodynamics Boltzmann Generators
Conclusion
References
12 Defining the Role of Chemical and Biological Data and Applying Algorithms
1 Introduction
2 Data Mining and Its Application in Healthcare
2.1 History of Databases and Data Mining
2.2 Work of Data Mining
2.3 Data Mining in Healthcare
2.4 Expanding Data Mining in Healthcare
2.5 Results of Comparative Analysis of Various Diseases in Healthcare
2.6 Advantages of Data Mining in Healthcare
2.7 Future of Data Mining in Healthcare
3 Bioinformatics and Databases Building and Skills By Applying Algorithms
3.1 Data Mining
3.2 Algorithms
3.3 Basic Data Mining Process
3.4 Association Rule Mining Algorithm
3.5 Need for Data Mining in Bioinformatics
4 Power of Algorithms in Biology
4.1 Difficulty of Algorithms in Computational Biology
A Classic Computational Challenge Is Multiple Sequence Alignment
Complexity and Artificial Intelligence
Heuristics
Computational Biology Using High-Performance Computing
5 Conclusion
References
13 Optimization and Quantification of Uncertainty in Machine Learning
1 Introduction
1.1 Aim and Scope of Chapter
1.2 Chapter Organization
1.3 Defining Uncertainty and Its Types
1.3.1 Two Types of Uncertainty in Machine Learning
1.4 Four Main Sources of Uncertainty
1.5 Classification of Uncertainty Based On Input Domain
2 Uncertainty Quantification (UQ)
2.1 Stochastic UQ in Deep Learning Models
2.1.1 Monte Carlo (MC) Dropout Method
2.1.2 Markov Chain Monte Carlo (MCMC)
2.1.3 Variational Inference (VI)
2.1.4 Bayesian Active Learning (BAL)
2.1.5 Bayes By Backprop (BBB)
2.1.6 Variational Autoencoders (VAEs)
2.1.7 Laplacian Approximations
2.1.8 Ensemble Techniques
2.2 Deterministic UQ Methods in Deep Learning
2.2.1 Single Deterministic Methods
2.2.2 Test Time Data Augmentation
3 Uncertainty Optimization
3.1.1 Stochastic Programming
3.1.2 Chance Constrained Optimization
3.1.3 Robust Optimization
4 Conclusion
References
Index