Research On The Association Between AD Candidate Gene Interaction And MRI Images Based On FP-growth Algorithm

As the most common type of dementia,Alzheimer’s disease(AD)has become the main cause of death.The inability to take care of oneself in life,the gradual decline of memory,and the deterioration of cognitive function are main characteristics of AD.In recent years,the number of people dying from Alzheimer’s disease has increased significantly due to the lack of effective treatment.When the patient’s symptoms are obvious,the AD process has entered an advanced stage and is irreversible.Therefore,it is of great significance to the early diagnosis and prevention of AD.There are other research areas combine neuroimaging and genomics to study the association between image phenotypes and gene mutations,and to explore how gene mutations affect the structure and feature of the brain.In this paper,based on the brain imaging genetics,the brain imaging exploration is developed using genome-wide genetic data and human brain imaging data from the same group of subjects.By exploring the relationship between genetic variation and imaging phenotype,the diffusion-weighted magnetic resonance imaging data is standardized into the MNI space,and the gray matter density map is extracted.The genotyping data using in this paper are downloaded from the ADNI database(ADNI-GO/2).The goal of the database is to measure the progress of early AD.After analyzing the list of AD candidate genes,the ±20Kbp SNPs of the 24 AD genes that meet the conditions are extracted.The frequent itemset mining algorithm is used to analyze the genome-wide association analysis results between the magnetic resonance imaging data and the genotyping data,and the combination of SNP frequent item sets with significant correlation is obtained.And the meaningful association rule of SNP frequent itemset combination is obtained since the selection algorithm also has the function of mining association rules.The genome-wide association analysis method is used to analyze the association between the genotype and MRI data from the Alzheimer’s Disease Neuroimaging Initiative(ADNI),and the frequent itemset mining algorithm is applied to process the results.Different from previous studies on the effect of single SNP on brain imaging,the aim of this article is seeking multiple SNP frequent item sets that can affect brain imaging by combining other SNPs and are easily missed for some reasons.We finally obtain twelve 2-SNP combinations,eight3-SNP combinations and one 4-SNP combination that meet the support conditions.To verify the biology meaning of these combinations,the software named Mango is applied to analyze the SNP frequent itemset combinations by Slice brain slice analysis and three-dimensional perspectives of the brain from different angles.In addition,by analyzing the significance of the association rules,4 more meaningful 3-SNP combinations were selected,mapped and analyzed.Finally,the Go-Term analysis was performed on the obtained 4-SNP combination and it is observed that the mapped gene sets {RTF2,CSTF1,CASS4} of the SNP frequent itemset combinations {rs6092321,rs7806,rs386274,rs6024860} affect the initiation,stability and termination of protein replication.This is significant for the early diagnosis and intervention of AD.