Pheripheral Blood Transcriptomic Disturbance Analysis And Enomic Rare Variant Association Analysis On Alzheimer’s Disease

Blood transcriptome has emerged as a potential resource for the discovery of biomarkers for Alzheimer’s disease(AD). However, the validity of blood transcriptome in the early diagnosis of AD has yet to be extensively tested. In this work, we analyzed published data on AD blood transcriptome and revealed the characteristic perturbation of cellular functional units, including upregulation of environmental responses and down-regulation of core metabolism. This characteristic perturbation was unique to AD based on the comparison with blood transcriptome from other neurological disorders and complex diseases. More importantly, similar perturbation was observed in both AD and mild cognitive impairment (MCI) groups. This perturbation pattern was further validated in our independent microarray experiment in a small Chinese cohort. What’s more, we discussed possible relationship between the alteration of AD blood transcriptome and AD pathology.Genomic variantions associated with AD has been extensively researched for decades. Now it is believed that variants involved in a polygenic disease may work interactively. It is difficult to elucide the genomic changes related with polygenic disease. Gene linkage study was mainly used for assessing the genetic components of Alzheimer disease before the advent of genome wide association study(GWAS). GWAS has greatly helped us to discover common SNPs associated with AD. Lambert et al found21SNPs related with AD through GWAS meta anaysis with74046samples. But, the results found by GWAS cann’t fully explain what we have observed in the genetics of AD. The missing heredity, what it is called, has led us to the "common disease-rare variant" hypothesis. With whole genome sequencing data released by ADNI(Alzheimer Disease Neuron Imaging), we found137AD associated rare variants genes with3different "gene-based" rare variant association testing methods. For a systematic filtration and interpretion of the candidate genes we found, a bayesian based scoring network was constructed and used to do further analysis of these rare variants related genes.