Neurodegenerative diseases (NDDs) are characterized by progressive neurological impairment caused by the accumulation of abnormal proteins and neuronal loss. The abnormal proteins apparently alter neuronal functions that lead to the disruption of synapses in neuronal sub-populations, neural circuitry and higher-order neural architectures within specific regions of the brain. Since neurological deficits are not necessarily associated with neuronal loss , it is very likely that neurodegenerative conditions might be caused by neuronal dysfunction. The challenge lies in understanding how aberrations in gene regulation, protein-protein interactions, and the consequent alterations in signaling and metabolic pathways results in neuronal dysfunction. The task is even more daunting considering the vast data that is being created by genomic and proteomic research in biological sciences. From the perspective of developing new therapies, important questions arise. How does one decipher signal-gene-protein interactions that drive neurodegeneration? How does one integrate interaction networks and clinical data to predict cognitive impairment and disease? To answer these questions, it will be necessary to rummage through information scattered across public-domain websites and research literature. Our objective is to gather and consolidate this information under one portal for NDDs and provide network tools to interrogate the data for identifying critical genes, determine pathways that are aberrant, create protein-protein interaction (PPI) networks to interpret disease mechanisms and perform qualitative and quantitative network analyses.
Previous instances of databases created to address specific problems underscores the importance of bringing together information under a common fold for analyses. The unifying principle for integrating this information was protein and gene interactions across species as in the case of BioGRID. BioGRID provides the interactions for Saccharomyces cerevisiae, Schizosaccharomyces pombe, Caenorhabditis elegans, Drosophila melanogaster, Mus musculus and Homo sapiens. Each interaction record in BioGRID is based on experimental evidence and is linked to the supporting publication. MINT on the other hand contains the molecular interactions experimentally verified and reported in peer-reviewed journals . Similarly, Reactome, a database of human pathways, was created with entries cross-referenced in a vocabulary associated with standard databases such as Uniprot, NCBI Entrez Gene, Ensembl, UCSC, HapMap, KEGG and primary research literature to PubMed. It describes the role of 5272 human proteins and 3504 macromolecular complexes in 3847 reactions organized into 1057 pathways . Sage Bionetworks, ELIXIR, Biomart and InterMine, have recognized the value of collecting, curating and categorizing data, and have undertaken the colossal task of creating infrastructure for the management of open source databases [5–7]. Many other database resources available online have also been developed to unify a class of data of interest .
Data related to neurodegenerative diseases has also grown exponentially with recent advances in high throughput genotyping techniques using microarrays. To enable interpretation of the insurmountable data, databases have been created to gather and rationalize the impact of mutations and protein-protein interactions on clinical manifestation of individual diseases. Some of the examples include the databases for Alzheimer’s disease (AD), amyotrophic lateral sclerosis (ALS) and Parkinson’s disease (PD). The AD database AlzGene, was developed to understand the genetic proclivity of AD and predict candidates for other complex genetic diseases . AlzGene catalogues all genetic association studies published in the field of AD. Meta-analyses results of polymorphisms with genotypes are publicly available at this site. Yang et al. have floated a database that contains experimentally confirmed substantianigra expressed sequence tags from healthy and PD patients. The database captures genetic variation, differential gene expression, gene-regulating elements, mitochondrial proteins, and pathways associated with PD-related genes. To integrate genetic and clinical information on ALS, Yoshida et al. developed a database that provides 180 unique variants identified in ALS patients along with the corresponding clinical data. These databases are useful to both experimentalists and theorists who wish to understand data pertaining to a single disease of their interest in relation to known information.
Many examples reported in the literature show that reconstruction and analyses of these networks has given a deeper understanding of disease mechanisms and strategies for therapeutic intervention. Goh et al. has shown that NDDs possess one of the most connected networks through a disease-gene network analysis. Understanding these networks is important because genes associated with a disease are not randomly positioned, but occur in clusters that are positively correlated with other similar diseases . The study of pathway-based genetic analysis in multiple sclerosis (MS) indicated that the understanding of biological mechanisms of disease pathogenesis and identification of drug targets may come from distant associations . Hwang et al. performed dynamic systems analyses to identify perturbations of cellular processes that were required for prion replication. PPI analysis of a network created by combining library and matrix yeast two-hybrid screens, led to the discovery that GIT1, a GTPase activating protein that modulates actin polymerization, spine morphology, and synapse formation in neuronal cells, enhanced the aggregation of the huntingtin protein. Further, through these analyses, they detected 6 new huntingtin interacting proteins of unknown function . Limviphuvadh et al. focused on protein–protein interaction networks associated with causative proteins of six neurodegenerative disorders. They investigated correlation among NDDs using domain characteristics and found that PD and HD showed highest correlation among them. However, the challenge lies in the development of a theoretical framework that will enable the organization of existing data, and permit the interrogation and interpretation of mechanisms causing disease.
It is in this light that we have created a database, NeuroDNet, that includes information about twelve neurodegenerative diseases - adrenomyeloneuropathy, Alzheimer disease, amyotrophic lateral sclerosis, ataxia-telangiectasia, dentatorubral-pallidoluysian atrophy, Friedreich ataxia, frontotemporal dementia, Huntington disease, Lewy body dementia, Parkinson disease, prion disease, progressive supranuclear palsy. It accounts for the interactions and regulation between signaling molecules, genes and proteins. This database is also the first of its kind, which enables the construction and analysis of NDDs through PPI, regulatory and Boolean networks. We also present the results of three case studies, which demonstrate the power of the analytical tools featured in NeuroDNet.