基于先验算法的AD基因与影像的关联研究

Research on the Association between AD Gene and Image Based on Apriori Algorithm

阿尔兹海默病(Alzheimer’s disease,AD)已成为导致死亡的主要原因之一。AD的早期诊断和干预成为大多数研究关注的焦点。本文针对多个SNP的相互关联,将影像表型和基因变异之间进行关联研究。首先处理ADNI-GO/2期的基因分型数据,提取位于24个AD候选基因中±20K bp的SNPs;其次处理1515名参与者的MRI图像,得到1515个182×218×182的voxel影像数据,对获得的voxel影像数据进行降维处理,得到1515个61×73×61大小的数据;最后综合SNP数据与voxel数据,进行vGWAS(voxel Genome-Wide Association Study)分析。将关联分析结果应用频繁项集挖掘算法中的先验算法,得到满足覆盖率40%的具有遗传效应且区域覆盖率最大的SNP集合{rs6092321,rs7806,rs386274},本文揭示多个SNP在脑功能结构中的交互作用,对阿尔兹海默症早期诊断有重要意义。

Alzheimer’s disease (AD) has become one of the main causes of death.Early diagnosis and intervention of AD have become the focus of most research.This article focuses on the correlation of multiple SNPs,and conducts a correlation study between image phenotype and gene variation.Firstly process the genotyping data of ADNI-GO/2 phase,extract the SNPs located at±20K bp in the 24 AD candidate genes;secondly process the MRI images of 1515 participants,and get 1515 voxel image data of 182×218×182,Perform dimensionality reduction processing on the obtained voxel image data,and get 1515 data with a size of 61×73×61;finally,synthesize the SNP data and voxel data to perform vGWAS (voxel Genome-Wide Association Study) analysis.The association analysis result is applied to the prior algorithm in the frequent itemset mining algorithm,and the SNP set {rs6092321,rs7806,rs386274} with the genetic effect and the largest area coverage meeting the coverage rate of 40% is obtained.The interaction in the structure is of great significance for the early diagnosis of Alzheimer’s disease.