Hidden Markov Models for Controlling False Discovery Rate in Genome-Wide Association Analysis
Genome-wide association studies (GWAS) have shown notable success in identifying susceptibility genetic variants of common and complex diseases. To date, the analytical methods of published GWAS have largely been limited to single single nucleotide polymorphism (SNP) or SNP–SNP pair analysis, coupled with multiplicity control using the Bonferroni procedure to control family wise error rate (FWER). However, since SNPs in typical GWAS are in linkage disequilibrium, simple Bonferonni correction is usually over conservative and therefore leads to a loss of efficiency. In addition, controlling FWER may be too stringent for GWAS where the number of SNPs to be tested is enormous. It is more desirable to control the false discovery rate (FDR). We introduce here a hidden Markov model (HMM)-based PLIS testing procedure for GWAS. It captures SNP dependency by an HMM, and based which, provides precise FDR control for identifying susceptibility loci.
Key wordsGenome-wide association SNP Hidden Markov model False discovery rate EM algorithm Multiple tests
- 2.Sabatti C, Service S, Freimer N (2003) False discovery rate in linkage and association genome screens for complex disorders. Genetics 164:829–833.Google Scholar
- 3.Benjamini Y, Hochberg Y (1995) Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society. Series B (Methodological) 57:289–300.Google Scholar
- 11.Weinberg W (1908) Über den Nachweis der Vererbung beim Menschen. Jahresh Wuertt Ver vaterl Natkd 64:368–382.Google Scholar
- 12.Rabiner LR (1989) A tutorial on hidden markov models and selected applications in speech recognition. In Proceedings of the IEEE, p.257–286.Google Scholar
- 14.Fisher RA (1932) Statistical Methods for Research Workers. Oliver & Boyd, EdinburghGoogle Scholar