Artificial intelligence-driven meta-analysis of brain gene expression identifies novel gene candidates and a role for mitochondria in Alzheimer’s Disease

Abstract

Microarrays have identified thousands of dysregulated genes in the brains of patients with Alzheimer's disease (AD); yet identifying the best gene candidates to both model and treat AD remains a challenge. To this end, we performed a meta-analysis of microarray data from the frontal cortex and cerebellum of AD patients, followed by an artificial-intelligence driven approach to identify the top AD gene candidates. In the frontal cortex, gene candidates included mitochondrial complex V subunits: ATP5J, ATP5L and ATP5H. In the cerebellum, the top candidate was SC5D, which is involved in cholesterol biosynthesis and catabolism. An important finding was that there was no overlap in dysregulated pathways between the frontal cortex and cerebellum, suggesting that pathophysiological mechanisms in AD are brain-region specific. Combined, these results have significant implications for future models and therapeutic strategies in AD.

Optimizing Alzheimer's disease prediction using the nomadic people algorithm

“…From thousands of attributes that are usually investigated, very few of these attributes display relevance to disease. So, must keep only the relevant features [25]. Studying the accurate selection of genes for the categorization process is aided by a gene profile.…”

“…As a result, identifying a minimal number of genes is crucial, known as informative genes that may be sufficient for acceptable classification. The most significant subset of genes, on the other hand, is frequently unknown [25], [26]. Filter and wrapper methods are two common gene selection techniques.…”

Artificial Intelligence and Circulating Cell-Free DNA Methylation Profiling: Mechanism and Detection of Alzheimer’s Disease

“…Systems biology techniques to include epigenetics, genomics, transcriptomics and metabolomics combined with both clinical and imaging data have been used to perform classification, biomarker identification, disease subtyping, disease progression prediction and drug repurposing [ 24 ]. With regard to transcriptomic studies and PM brain studies, AI analysis identifies candidate genes for AD [ 25 ]. Most of the reported AI based studies used invasive biomatrices such as PM brain tissue and the expensive imaging modalities.…”