Disease‐Concordant Twins Empower Genetic Association Studies. (23rd December 2016)

• Record Type: Journal Article
• Authors: Tan, Qihua; Li, Weilong; Vandin, Fabio
• Is Part Of: Annals of human genetics. Volume 81:Number 1(2017:Jan.)
• Page Start: 20
• Page End: 26
• Publication Date: 2016-12-23
• Subjects: Twins -- disease concordance -- power -- simulation -- genetic association; Human genetics -- Periodicals; 599.935
• DOI: 10.1111/ahg.12181
• Languages: English
• ISSNs: 0003-4800

Abstract

Summary: Genome‐wide association studies with moderate sample sizes are underpowered, especially when testing SNP alleles with low allele counts, a situation that may lead to high frequency of false‐positive results and lack of replication in independent studies. Related individuals, such as twin pairs concordant for a disease, should confer increased power in genetic association analysis because of their genetic relatedness. We conducted a computer simulation study to explore the power advantage of the disease‐concordant twin design, which uses singletons from disease‐concordant twin pairs as cases and ordinary healthy samples as controls. We examined the power gain of the twin‐based design for various scenarios (i.e., cases from monozygotic and dizygotic twin pairs concordant for a disease) and compared the power with the ordinary case‐control design with cases collected from the unrelated patient population. Simulation was done by assigning various allele frequencies and allelic relative risks for different mode of genetic inheritance. In general, for achieving a power estimate of 80%, the sample sizes needed for dizygotic and monozygotic twin cases were one half and one fourth of the sample size of an ordinary case‐control design, with variations depending on genetic mode. Importantly, the enriched power for dizygotic twins also applies to disease‐concordant sibling pairs, which largely extends the application of the concordant twin design. Overall, our simulation revealed a high value of disease‐concordant twins in genetic association studies and encourages the use of genetically related individuals for highly efficiently identifying both common and rare genetic variants underlying human complex diseases without increasing laboratory cost.