Recent Advances in Biological Network Analysis: Comparative Network Analysis and Network Module Detection

✍ Scribed by Byung-Jun Yoon (editor), Xiaoning Qian (editor)

• Publisher: Springer
• Year: 2021
• Tongue: English
• Leaves: 220
• Category: Library

✦ Synopsis

This book reviews recent advances in the emerging field of computational network biology with special emphasis on comparative network analysis and network module detection. The chapters in this volume are contributed by leading international researchers in computational network biology and offer in-depth insight on the latest techniques in network alignment, network clustering, and network module detection. Chapters discuss the advantages of the respective techniques and present the current challenges and open problems in the field.

Recent Advances in Biological Network Analysis: Comparative Network Analysis and Network Module Detection will serve as a great resource for graduate students, academics, and researchers who are currently working in areas relevant to computational network biology or wish to learn more about the field. Data scientists whose work involves the analysis of graphs, networks, and other types of data with topological structure or relations can also benefit from the book's insights.

✦ Table of Contents

Preface
Contents
Contributors
Part I Comparative Network Analysis
1 Global Alignment of PPI Networks
1.1 Introduction
1.2 Preliminary Definitions
1.3 Balanced Global One-to-One Network Alignment
1.3.1 Problem Definition
1.3.2 Computational Complexity
1.3.3 Algorithms
1.4 Constrained Global One-to-One Network Alignment
1.4.1 Problem Definition
1.4.2 Computational Complexity
1.4.3 Algorithms
1.5 Global Many-to-Many Network Alignment
1.5.1 Problem Definition
1.5.2 Computational Complexity
1.5.3 Algorithms
1.6 Network Alignment Evaluation Criteria
1.6.1 Procedures and Metrics with No Biological Connotation
1.6.2 Procedures and Metrics with Biological Connotation
1.7 Discussion
References
2 Effective Random Walk Models for Comparative NetworkAnalysis
2.1 Introduction
2.2 Methods
2.2.1 Problem Formulation
2.2.2 Semi-Markov Random Walk Model
2.2.3 Context-Sensitive Random Walk Model
2.2.4 CUFID Model
2.2.5 PPI Network Alignment Algorithms Using the Random Walk Models
2.3 Results
2.3.1 Datasets
2.3.2 Performance Metrics
2.3.3 Performance Assessment Through Synthetic PPI Networks
2.3.4 Performance Assessment Through Real PPI Network
2.4 Discussion
Notes
References
3 Computational Methods for Protein–Protein Interaction Network Alignment
3.1 General
3.1.1 Protein–Protein Interaction Network
3.1.2 Protein–Protein Interaction Network Alignment
3.1.3 Experimental Datasets
3.1.4 Evaluation of Alignment Quality
3.1.4.1 Functional Metrics
3.1.4.2 Topological Metrics
3.2 Alignment Methods
3.2.1 HubAlign: PPI Network Alignment by Topological Importance
3.2.1.1 Introduction
3.2.1.2 Method Details of HubAlign
3.2.1.3 Performance of HubAlign
3.2.2 ModuleAlign: Module-based Global Alignment of PPI Networks
3.2.2.1 Introduction
3.2.2.2 Method Details of ModuleAlign
3.2.2.3 Performance of ModuleAlign
3.2.3 ConvexAlign: Multiple Network Alignment via Convex Optimization
3.2.3.1 Introduction
3.2.3.2 Method Details of ConvexAlign
3.2.3.3 Performance of ConvexAlign
3.3 Conclusion
References
4 BioFabric Visualization of Network Alignments
4.1 Background
4.1.1 Our Contributions
4.2 Methodology
4.2.1 Overview and Nomenclature
4.2.2 Node and Link Groupings
4.2.3 Network Merge
4.2.4 New Metrics
4.2.5 Node and Link Group Layout Algorithm
4.2.6 Alignment Cycle Layout for Self-alignments
4.2.7 Interface with BioFabric
4.3 Results
4.3.1 Case Study I: Simple Network Comparison
4.3.2 Case Study II: Visualizing Common Topological Measures
4.3.3 Case Study III: Visualizing Performance of Objective Functions
4.3.4 Case Study IV: Finding Protein Cluster Misalignments
4.4 Discussion
4.4.1 Limitations
4.4.2 Future Work
4.5 Conclusions
Appendix
Link and Node Groups with Blue Nodes
Jaccard Similarity with Blue Nodes
Creation of the Correct Network Alignment
Detailed Description of the Node Assignment Algorithm for the Node and Link Group Layout
Full Table of All Alignment Scores for Mixtures of Importance and Symmetric Substructure Score
Table of Node Group Sizes for Case III
Percentage of Purple Nodes Without and with Incident pRr Edges Between Correct and Mixed Alignments
Alignment Cycle Layout with Blue Nodes
The Four-Cluster Misalignment
References
Part II Network Module Detection
5 Motifs in Biological Networks
5.1 Introduction and Background
5.2 Definitions and Notation
5.3 Existing Algorithms
5.3.1 Number of Input Networks
5.3.1.1 Single Input Network
5.3.1.2 Multiple Input Networks
5.3.2 Graph Labeling
5.3.3 Frequency Formulation
5.3.4 Directed and Undirected Graphs
5.3.5 Network Model
5.3.5.1 Probabilistic Model
5.3.5.2 Dynamic Model
5.4 Conclusion
References
6 Module Identification of Biological Networks via Graph Partition
6.1 Biological Networks
6.1.1 Protein Interaction Networks
6.1.2 Gene Co-expression Networks
6.1.3 Cell–Cell Similarity Networks
6.2 What Is the Definition of a Module?
6.2.1 Modularity
6.2.1.1 Adjacency Matrix of the Null Model P
6.2.1.2 The Resolution Limitation
6.2.1.3 Multiresolution Modularity
6.2.2 Conductance
6.2.2.1 Normalized Cut
6.2.2.2 Block Modeling
6.3 Graph Partition Algorithms for Module Identification
6.3.1 Algorithms for Modularity Maximization
6.3.1.1 Spectral Method
6.3.1.2 Louvain Algorithm
6.3.2 PageRank-Nibble Algorithm for ConductanceMinimization
6.3.2.1 Unweighted Network
6.3.2.2 Weighted Network
6.3.2.3 Markov Cluster Algorithm
6.3.3 Algorithms for the Normalized Cut Problem
6.3.3.1 Spectral Method
6.3.3.2 Semidefinite Programming Method
6.3.4 Algorithms for Block Modeling
6.3.4.1 Simulated Annealing
6.3.4.2 Subgradient-Based Optimization Algorithm
6.4 Conclusion
References
7 Network Module Detection to Decipher Heterogeneity of Cancer Mutations
7.1 Introduction
7.2 Identifying Cancer Driver Genes and Mutations
7.3 Phenotype-Associated Mutations
7.4 Interactions Between Modules
7.5 Conclusion
References
Part III Network-Based Omics Data Analysis
8 Integrated Network-Based Computational Analysis for Drug Development
8.1 Introduction
8.2 Prediction of Therapeutic Targets
8.3 Prediction of Drug Repositioning
8.4 Prediction of Drug Sensitivity
8.5 Prediction of Drug–Drug Interaction
8.6 Prediction of Hormone–Drug Interaction
8.7 Prediction of Therapeutic Effects of Natural Products
8.8 Conclusion
References
9 Network Propagation for the Analysis of Multi-omics Data
9.1 What Is Network Propagation
9.1.1 How Network Propagation Works
9.1.2 Properties of Network Propagation
9.2 Genome Data Analysis at DNA Level
9.3 Detecting Rare Somatic Mutations Using Network Propagation
9.3.1 Disease Subtype Classification Using Supervised Network Propagation
9.3.1.1 How Supervised Network Propagation Works
9.3.1.2 Where the Supervised Network Propagation Is Applied
9.4 Transcriptome Data Analysis at RNA Level
9.4.1 Gene Prioritization
9.4.2 PropaNet: TF Prioritization and Subnetwork Construction on Time-Series Gene Expression Data
9.4.2.1 How Network Propagation Is Utilized Within PropaNet
9.4.3 Differentially Expressed Gene (DEG) Detection
9.4.3.1 How Network Propagation Is Used to Extract a Feature for Machine Learning
9.4.4 Hypothesis Testing
9.4.4.1 How Network Propagation Is Utilized to Evaluate Hypotheses
9.4.5 Comparing Multiple Experiments
9.4.5.1 How Network Propagation Is Utilized in Venn-diaNet
9.5 Transcriptome Data Analysis at Pathway Level
9.5.1 Disease Subtype Classification with Pathway Information
9.5.1.1 What Is the Attention-Based Ensemble Model of Graph Convolutional Networks
9.5.1.2 How Network Propagation Is Utilized in the Attention-Based Ensemble Model of Graph Convolutional Networks
9.6 Conclusion
References