Machine Learning Methods for Multi-Omics Data Integration

The advancement of biomedical engineering has enabled the generation of multi-omics data by developing high-throughput technologies, such as next-generation sequencing, mass spectrometry, and microarrays. Large-scale data sets for multiple omics platforms, including genomics, transcriptomics, proteomics, and metabolomics, have become more accessible and cost-effective over time. Integrating multi-omics data has become increasingly important in many research fields, such as bioinformatics, genomics, and systems biology. This integration allows researchers to understand complex interactions between biological molecules and pathways. It enables us to comprehensively understand complex biological systems, leading to new insights into disease mechanisms, drug discovery, and personalized medicine. Still, integrating various heterogeneous data types into a single learning model also comes with challenges. In this regard, learning algorithms have been vital in analyzing and integrating these large-scale heterogeneous data sets into one learning model. This book overviews the latest multi-omics technologies, machine learning techniques for data integration, and multi-omics databases for validation. It covers different types of learning for supervised and unsupervised learning techniques, including standard classifiers, deep learning, tensor factorization, ensemble learning, and clustering, among others. The book categorizes different levels of integrations, ranging from early, middle, or late-stage among multi-view models. The underlying models target different objectives, such as knowledge discovery, pattern recognition, disease-related biomarkers, and validation tools for multi-omics data. Finally, the book emphasizes practical applications and case studies, making it an essential resource for researchers and practitioners looking to apply machine learning to their multi-omics data sets. The book covers data preprocessing, feature selection, and model evaluation, providing readers with a practical guide to implementing machine learning techniques on various multi-omics data sets. Contents 5 Introduction to Multiomics Technology 7 1 Genomics 7 2 Transcriptomics 9 3 Proteomics 11 4 Foodomics 12 5 Metabolomics 13 6 Epigenomics 14 7 Summary 15 References 16 Machine Learning from Multi-omics: Applications and DataIntegration 18 1 Introduction 18 2 Multi-omics as Cancer Indicators 19 3 Multi-omics Epigenetic Alterations of Alzheimer's 20 4 Multi-omics Applications in Mental Health and Psychiatric Disorders 20 5 Cardiovascular Disease 21 6 Machine Learning Applications to Multi-omics Data 21 7 Data Integration Strategies 22 References 25 Machine Learning Approaches for Multi-omics Data Integration in Medicine 27 1 Introduction 27 2 Main Objectives of Multi-omic Data Integration Studies 30 2.1 Diagnosis and Prognosis 30 2.2 Identification of the Subtype 30 2.3 Discovering Molecular Patterns of Disease 30 2.4 Predicting the Effects of a Drug at the Molecular Level 31 2.5 Comprehension of the Regulatory Processes 31 3 Multi-omics Integration Strategies 31 3.1 Early Integration 32 3.2 Mixed Integration 33 3.3 Intermediate Integration 33 3.4 Late Integration 34 3.5 Hierarchical Integration 34 4 Machine Learning Approaches Used in Multiomics Integration 34 4.1 Data Integration Analysis for Biomarker Discovery Using Latent Components (DIABLO) 34 4.2 Multi-omics Factor Analysis (MOFA) 35 4.3 Sparse Canonical Correlation Analysis (sCCA) 35 4.4 Multi-omics Late Integration (MOLI) 36 4.5 Cancer Drug Response Prediction Using a Recommender System (CaDRReS) 36 4.6 Heterogeneous Network-Based Method for Drug Response Prediction (HNMDRP) 36 4.7 Multiple Pairwise Kernels for Drug Bioactivity Prediction (pairwiseMKL) 37 4.8 iCluster, iClusterPlus, and iClusterBayes 37 4.9 moCluster 37 4.10 Similarity Network Fusion (SNF) 38 4.11 NEighborhood Based Multi-omics Clustering (NEMO) 38 4.12 Random Walk with Restart for Multi-dimensional Data Fusion (RWRF) and Random Walk with Restart and Neighbor Information-Based Multi-dimensional Data Fusion (RWRNF) 39 5 Conclusion 39 References 40 Multimodal Methods for Knowledge Discovery from Bulk and Single-Cell Multi-Omics Data 43 1 Introduction 43 2 Description of Various Omics Datasets 45 2.1 ChIP-seq 45 2.2 ATAC-seq 48 2.3 Hi-C 49 2.4 Mass Spectrometry for Proteomics 49 2.5 Single-Cell Multi-Omic Profiling 50 3 Multimodal Methods for Dimensionality Reduction and Clustering 50 3.1 Non-negative Matrix Factorization 52 3.2 Tensor Decomposition 53 3.3 Multi-View Relational Learning 54 3.4 Canonical Correlation Analysis 54 3.5 Deep Learning Methods for Multimodal Dimension Reduction and Clustering 55 3.6 Evaluating and Visualizing Single-Cell Embeddings 56 4 Multimodal Methods for Inferring Gene Regulatory Networks from Bulk and Single-Cell Omics Data 56 4.1 Multiple Regression 57 4.2 Correlation and Mutual Information 59 4.3 Ordinary Differential Equation 61 5 Multi-Modal Network Inference of Gene Regulations 62 5.1 Bayesian Network Inference 62 5.2 Static Boolean Regulatory Network Inference 64 5.3 Dynamic Regulatory Network Inference 64 6 Multimodal Methods for Biomarker Identification 66 6.1 Ensemble Learning Based Multi-Omic Biomarker Identification 66 6.2 Deep Neural Network Based Multi-Omic Biomarker Identification 67 7 Closing Remarks and Perspectives 68 References 72 Negative Sample Selection for miRNA-Disease Association Prediction Models 79 1 Introduction 79 2 Methods 81 2.1 Obtain the Feature Representations for Each miRNA-Disease Sample 81 2.2 Train the Deep Autoencoder with All the Verified Samples 82 2.3 Sort All the Unknown Samples by the Deep Autoencoder 84 3 Result 84 3.1 Database 84 3.2 Evaluation Methods 85 3.3 Experiment Setting and Overfitting Analyzing 85 3.4 Reconstruct Error Data Analysis on Well Trained DAE-N Model 87 3.5 Compared Methods 88 3.6 Effectiveness of DAE-N on Cross-Validation 89 3.7 Effectiveness of DAE-N on Independent Dataset Evaluation 91 4 Conclusion 92 References 93 Prediction and Analysis of Key Genes in Prostate Cancer via MRMR Enhanced Similarity Preserving Criteria and Pathway Enrichment Methods 95 Acronyms 95 1 Introduction 96 2 Literature Review 97 2.1 Feature Selection Methods 98 2.1.1 Fisher Score 99 2.1.2 Laplace Score 99 2.1.3 ReliefF Criteria 99 2.1.4 Unified Framework for Similarity Based Methods 100 2.1.5 MRMR 101 2.2 Description of Classifiers 101 2.3 Pathway Enrichment Analysis 102 3 Methods 103 3.1 Data Source and Data Type 104 3.2 Data Preparation 104 3.3 Experiment Design 104 3.4 Feature Selection 105 3.4.1 The Problem 105 3.4.2 The Algorithm 105 3.4.3 Classification 106 3.5 Measures for Performance Evaluation 107 3.6 Enrichment Analysis of Key Pathways and Core Genes 107 4 Results and Discussion 108 4.1 Identification of Key Genes Related to PCa 113 4.2 GO and KEGG Pathway Enrichment Analyses 113 4.3 Discussion 115 5 Conclusions 116 References 117 Graph-Based Machine Learning Approaches for Pangenomics 120 1 Introduction 120 2 Methods 122 2.1 Frequented Regions 122 2.2 Data 124 2.3 Genome-Wide Association Study 124 2.4 Machine Learning Models 125 2.5 Experimental Setup 126 3 Results 128 3.1 Phenotypic Prediction 128 3.2 FRs and Annotations 130 4 Conclusion 132 References 133 Multiomics-Based Tensor Decomposition for Characterizing Breast Cancer Heterogeneity 136 1 Breast Cancer Inter-Tumor Heterogeneity 136 1.1 Morphological and Histopathologic Heterogeneity 137 1.2 Biomarker Heterogeneity 138 1.3 Genetic Heterogeneity and Breast Cancer Subtyping Schemes 139 2 Breast Cancer Multiomics Data 140 2.1 Genomic Level: CNVs 140 2.2 Transcriptomic Level: Gene Expression 141 2.3 Epigenomic Level: DNA Methylation 141 3 Tensor-Based Multiomics Integration and Factorization 142 3.1 Tensor 142 3.2 Tensor Decomposition Algorithms 143 4 Applications 145 4.1 Breast Cancer Subtyping 145 4.2 Survival Prediction 146 4.3 Gene Set Enrichment Analysis 147 5 Conclusions 149 References 150 Multi-Omics Databases 154 Acronyms 154 1 Introduction 155 2 Literature Review 156 3 Multi-Omics Data Resources 157 3.1 Data Repositories 157 3.2 BioBanks 157 4 Multi-Omics Databases and Tools 158 5 Multi-Omics Main Technologies 161 6 Fields of Multi-Omics Technologies 163 7 Conclusion 166 References 167 Index 170