基于全基因组的遗传性疾病和性状的关联位点分析

研究发现,通过全基因组关联分析,找出与疾病相关的位点或基因,对于人们防治遗传病,具有重要意义.首先,考虑固定效应(SNP位点)和随机效应(人群中的群体结构和亲缘关系),建立了混合线性模型,并且利用基于FDR标准的BH法对多重检验的P值进行校正,找出最有可能的致病位点.其次,利用Fisher的P值组合方法,将基因所包含的所有SNP位点组合,找出与疾病最可能相关的基因.由于遗传疾病可能与基因所包含的位点的子集关联,我们参考已有的ARTP模型,对模型进行了改进.最后,建立多表型联合模型MultiPhen找出与10个性状有关联的位点.

Identification of SNP or Genes Associated with Genetic Diseases and Traits Based on GWAS

In recent years,genome-wide association studies (GWAS) have identified diseaseassociated genes and single nucleotide polymorphisms (SNPs),which is of great significance for the prevention and treatment of genetic disorders Firstly,we built a mixed linear model taking fixed effects (SNPs and population structure) and random effects into consideration.We identified pathogenic SNPs by adopting the BH step-up procedure to control the false discovery rate (FDR).Secondly,based on the first step,several genes were identified by combining Pvalues of all SNPs within one gene with Fisher's method.Given that diseases may be associated with the subset of SNPs includedin one gene,we improved our model with reference to the existing ARTP model.Finally,we tested the association between the linear combination of 10 phenotypes and the genotypes at each SNP by utilizing a joint model of multiple phenotypes(MultiPhen)