Identifying Cryptic Relationships
Cryptic relationships such as first-degree relatives often appear in studies that collect population samples, including genome-wide association studies (GWAS) and next-generation sequencing (NGS) analyses. Cryptic relatedness not only increases type 1 error rate of association tests but also affects other analytical aspects of GWAS and NGS such as population stratification via principal component analysis. Here, we discuss three effective methods, as implemented in PREST, PLINK, and KING, to detect and correct for the problem of cryptic relatedness using high-throughput SNP data collected from GWAS and NGS experiments. We provide the analytical and practical details involved using three application examples.
Key wordsCryptic relatedness Pedigree error Relationship estimation IBD IBS IIS Kinship coefficient Likelihood EM algorithm Method-of-moments Software PREST PREST-plus PLINK KING GWAS Sequencing
- 5.Dimitromanolakis A, Paterson AD, Sun L (2009) Accurate IBD inference identifies cryptic relatedness in 9 hapmap populations. Abstract no. 1768 presented at the annual meeting of the American Society of Human Genetics.Google Scholar
- 9.Chen WM, Manichaikul A, Rich SS (2016) KING 2.0: relationship inference and integrated analysis in one million samples. Abstract #365T presented at the American Society of Human Genetics annual meetingGoogle Scholar
- 11.Begleiter H, Reich T, Nurnberger JJ, Li TK, Conneally PM, Edenberg H, Crowe R, Kuperman S, Schuckit M, Bloom F, Hesselbrock V, Porjesz B, Cloninger CR, Rice J, Goate A (1999) Description of the genetic analysis workshop 11 collaborative study on the genetics of alcoholism. Genet Epidemiol 17(Suppl 1):S25–S30CrossRefPubMedGoogle Scholar
- 12.Antoni G, Morange P, Luo Y, Saut N, Burgos G, Heath S, Germain M, Biron-Andreani C, Schved J, Pernod G, Galan P, Zelenika D, Alessi M, Drouet L, Visvikis-Siest S, Wells P, Lathrop M, Emmerich J, Tregouet D, Gagnon F (2010) A multi-stage multi-design strategy provides strong evidence that the bai3 locus is associated with early-onset venous thromboembolism. J Thromb Haemost 8(12). doi: 10.1111/j.1538-7836.2010.04092.x