Abstract:Alzheimer's disease (AD) is an irreversible neurological disorder characterized by insidious onset. Identifying potential markers in its emergence and progression is crucial for early diagnosis and treatment. Imaging genetics typically merges genetic variables with multiple imaging parameters, employing various association analysis algorithms to investigate the links between pathological phenotypes and genetic variations, and to unearth molecular-level insights from brain images. However, most existing imaging genetics algorithms based on sparse learning assume a linear relationship between genetic factors and brain functions, limiting their ability to discern complex nonlinear correlation patterns and resulting in reduced accuracy. To address these issues, we propose a novel nonlinear imaging genetic association analysis method, Deep Self-Reconstruction-based Adaptive Sparse Multi-view Deep Generalized Canonical Correlation Analysis (DSR-AdaSMDGCCA). This approach facilitates joint learning of the nonlinear relationships between pathological phenotypes and genetic variations by integrating three different types of data: structural magnetic resonance imaging (sMRI), single-nucleotide polymorphism (SNP), and gene expression data. By incorporating nonlinear transformations in DGCCA, our model effectively uncovers nonlinear associations across multiple data types. Additionally, the DSR algorithm clusters samples with identical labels, incorporating label information into the nonlinear feature extraction process and thus enhancing the performance of association analysis. The application of the DSR-AdaSMDGCCA algorithm on real data sets identified several AD risk regions (such as the hippocampus, parahippocampus, and fusiform gyrus) and risk genes (including VSIG4, NEDD4L, and PINK1), achieving maximum classification accuracy with the fewest selected features compared to baseline algorithms. Molecular biology enrichment analysis revealed that the pathways enriched by these top genes are intimately linked to AD progression, affirming that our algorithm not only improves correlation analysis performance but also identifies biologically significant markers.

Joint Association Discovery and Diagnosis of Alzheimer's Disease by Supervised Heterogeneous Multiview Learning.

Supervised Heterogeneous Multiview Learning for Joint Association Study and Disease Diagnosis

Sparse Bayesian Multiview Learning for Simultaneous Association Discovery and Diagnosis of Alzheimer's Disease.

Deep joint learning diagnosis of Alzheimer's disease based on multimodal feature fusion

Joint Learning Framework of Cross-Modal Synthesis and Diagnosis for Alzheimer's Disease by Mining Underlying Shared Modality Information.

Exploration of Imaging Genetic Biomarkers of Alzheimer's Disease Based on a Machine Learning Method

Joint learning framework of cross-modal synthesis and diagnosis for Alzheimer’s disease by mining underlying shared modality information

Joint Coupled-Feature Representation and Coupled Boosting for AD Diagnosis

Multi-Modal Data Analysis for Alzheimer’s Disease Diagnosis: An Ensemble Model Using Imagery and Genetic Features

Diagnosis of Alzheimer's Disease Using View-Aligned Hypergraph Learning with Incomplete Multi-modality Data

Application of orthogonal sparse joint non-negative matrix factorization based on connectivity in Alzheimer's disease research

Exploring Imaging Genetic Markers of Alzheimer's Disease Based on a Novel Nonlinear Correlation Analysis Algorithm

Diagnosis-Guided Deep Subspace Clustering Association Study for Pathogenetic Markers Identification of Alzheimer's Disease Based on Comparative Atlases

Multivariate Deep Learning Classification of Alzheimer’s Disease Based on Hierarchical Partner Matching Independent Component Analysis

A Novel Longitudinal Phenotype-Genotype Association Study Based on Deep Feature Extraction and Hypergraph Models for Alzheimer's Disease

Joint analysis of multivariate longitudinal, imaging, and time-to-event data

Sparse Learning and Hybrid Probabilistic Oversampling for Alzheimer’s Disease Diagnosis

3D Multimodal Fusion Network With Disease-Induced Joint Learning for Early Alzheimer's Disease Diagnosis

Multi-modal imaging genetics data fusion by deep auto-encoder and self-representation network for Alzheimer's disease diagnosis and biomarkers extraction

Integration of Imaging Genomics Data for the Study of Alzheimer's Disease Using Joint-Connectivity-Based Sparse Nonnegative Matrix Factorization

An explainable two-dimensional single model deep learning approach for Alzheimer's disease diagnosis and brain atrophy localization