Network Analysis for Complex Neurodegenerative Diseases

Abstract

Purpose of Review Biomedicine is witnessing a paradigm shift in the way complex disorders are investigated. In particular, the need for big data interpretation has led to the development of pipelines that require the cooperation of different fields of expertise, including medicine, functional biology, informatics, mathematics and systems biology. This review sits at the crossroad of different disciplines and surveys the recent developments in the use of graph theory (in the form of network analysis) to interpret large and different datasets in the context of complex neurodegenerative diseases. It aims at a professional audience with different backgrounds. Recent Findings Biomedicine has entered the era of big data, and this is actively changing the way we approach and perform research. The increase in size and power of biomedical studies has led to the establishment of multi-centre, international working groups coordinating open access platforms for data generation, storage and analysis. Particularly, pipelines for data interpretation are under development, and network analysis is gaining momentum since it represents a versatile approach to study complex systems made of interconnected multiple players. Summary We will describe the era of big data in biomedicine and survey the major freely accessible multi-omics datasets. We will then introduce the principles of graph theory and provide examples of network analysis applied to the interpretation of complex neurodegenerative disorders. Keywords Complex neurodegeneration. Network analysis. Protein-protein interactions. GWAS loci. Omics data. Data integration. Gene co-expression This article is part of the Topical Collection on Bioinformatics

Integrating protein networks and machine learning for disease stratification in the Hereditary Spastic Paraplegias

“…Network-based approaches have been increasingly used to study complex human diseases, such as neurodegenerative diseases and cancer (Manzoni et al ., 2020). The Hereditary Spastic Paraplegias (HSPs) are neurodegenerative diseases with considerable genetic and clinical heterogeneity (Boutry et al ., 2019; Faber et al ., 2017), rendering them particularly interesting to study using a protein-protein interaction network (PPIN) approach.…”

“…In contrast, systems biology approaches such as protein-protein interaction (PPI) network (PPIN) analyses provide tools to evaluate the entirety of known genes/proteins involved in a disease collectively through a holistic approach (Koh et al, 2012). The connections within the PPIN can be subjected to mathematical analysis to gain insight into the global relationships among potential contributors to the disease process, thus creating an in silico model system to investigate the molecular mechanisms and generate hypotheses to further support functional research and disease modelling (Manzoni et al, 2020). This paper describes the first study in which PPINs are created solely based on experimentally validated human PPIs of HSP genes, and are applied to the investigation of HSP pathogenesis to identify global mechanisms, as well as individual processes involved in subtypes of disease following stratification based on the association of specific HSP genes with particular clinical features.…”

“…In contrast, systems biology approaches such as protein-protein interaction (PPI) network (PPIN) analyses provide tools to evaluate the entirety of known genes/proteins involved in a disease collectively through a holistic approach (Koh et al , 2012). The connections within the PPIN can be subjected to mathematical analysis to gain insight into the global relationships among potential contributors to the disease process, thus creating an in silico model system to investigate the molecular mechanisms and generate hypotheses to further support functional research and disease modelling (Manzoni et al ., 2020).…”

Common Factors in Neurodegeneration: A Meta-Study Revealing Shared Patterns on a Multi-Omics Scale

“…To address the challenges of such complicated diseases, the whole field of biomedicine is changing towards creating and facilitating a variety of databases and analysis pipelines for separate omics layers and multi-omics integration [ 285 ]. Many of these pipelines are mainly data-driven and enable clustering and supervised machine learning techniques to find essential patterns of features contributing to the identification of, for example, proteins that are associated with NDDs [ 286 ], or to reveal cross-talk patterns in multi-omics data [ 287 ].…”

Multi-Layer Picture of Neurodegenerative Diseases: Lessons from the Use of Big Data through Artificial Intelligence

“…Notably, the heterogeneity of ND is a key confounding factor that complicates the understanding of disease mechanisms and the identification of treatments. Case-control cohorts often include multiple phenotypes on distinct disease trajectories or rely on models that only account for a few features of the central nervous system at a time, which has been reductive for complex diseases [7][8][9]. Alzheimer's (AD) and Parkinson's (PD) diseases are two of the most frequent and heterogeneous pathologies among all the complex neurodegenerative proteinopathies, affecting 24 and 6.1 million people worldwide, respectively [3,7,10].…”