A probabilistic method for identifying rare variants underlying complex traits

BMC Genomics

Proceedings

Jiayin Wang¹  Jin Zhang¹  Zhongmeng Zhao²  Aiyuan Yang²  Zhi Cao² 

[1] Computer Science and Engineering Department, University of Connecticut, 06269-2155, Storrs, Connecticut, USA;Department of Computer Science and Technology, Xi'an Jiaotong University, Xi'an, P.R.China;

关键词: Minor Allele Frequency;    Rare Variant;    Causal Variant;    Markov Random Field;    Background Region;   

DOI  :  10.1186/1471-2164-14-S1-S11

来源: Springer

【 摘 要 】

BackgroundIdentifying the genetic variants that contribute to disease susceptibilities is important both for developing methodologies and for studying complex diseases in molecular biology. It has been demonstrated that the spectrum of minor allelic frequencies (MAFs) of risk genetic variants ranges from common to rare. Although association studies are shifting to incorporate rare variants (RVs) affecting complex traits, existing approaches do not show a high degree of success, and more efforts should be considered.ResultsIn this article, we focus on detecting associations between multiple rare variants and traits. Similar to RareCover, a widely used approach, we assume that variants located close to each other tend to have similar impacts on traits. Therefore, we introduce elevated regions and background regions, where the elevated regions are considered to have a higher chance of harboring causal variants. We propose a hidden Markov random field (HMRF) model to select a set of rare variants that potentially underlie the phenotype, and then, a statistical test is applied. Thus, the association analysis can be achieved without pre-selection by experts. In our model, each variant has two hidden states that represent the causal/non-causal status and the region status. In addition, two Bayesian processes are used to compare and estimate the genotype, phenotype and model parameters. We compare our approach to the three current methods using different types of datasets, and though these are simulation experiments, our approach has higher statistical power than the other methods. The software package, RareProb and the simulation datasets are available at: http://www.engr.uconn.edu/~jiw09003.

【 授权许可】

CC BY   
© Wang et al.; licensee BioMed Central Ltd. 2013