In silico identification of new genetic variations as potential risk factors for Alzheimer's disease in a microarray-oriented simulation.

Genomic and proteomic studies of neurodegenerative disorders require complementary approaches to integrate the massive amount of data generated in high throughput experimental procedures. We propose a Bioinformatics pipeline in which expression studies guide the selection of candidate genes that should be screened for potential new genetic variations from a public expressed site tags (ESTs) database. Motivated by the former interest of our group in genetic polymorphisms involved with the immune system, we selected five genes from a previous expression microarrays study of hippocampal cornu ammonis (CA1) area of Alzheimer's Disease subjects (AD). The CLCbio Workbench Combined version 3.6.2. was initially used to build ESTs and mRNA files retrieved respectively from the Goldenpath of University of California Santa Cruz (UCSC) and National Center for Biotechnology Information (NCBI) databases and latter to perform multiple batches of Smith-Waterman alignments. A total of 116 ESTs sequences were selected after proper stringent parameters were applied to the first set of mismatches. The annotation revealed various classes of variations, most of them deletions (176). Amongst this specific group, some were frameshift deletions (35) and the virtual translation of a few others (5) were predicted to induce no change other than a single aminoacid removal, with no subsequent repercussions at the protein sequence. In addition, the analysis identified transitions (three), transversions (52), synonymous (41), non-synonymous (12), and deletions in 36 ESTs located in Untranslated Regions -UTRs (Supplementary data). Deletions are often associated to major genetics syndromes with dysmorphic features. However, various recent studies show that common microdeletions might be highly associated with common neuropsychiatric disorders such as schizophrenia, autism, mental retardation, or even in various ethnicities, detected in whole genome sequencing experiments. A virtual validation confirmed that some of the variations identified were previously reported and confirmed in DNA samples, showing that this method is a feasible way to detect genetic variations that merit further exploration in AD genetic risk factor association studies.