Abstract
Genome-wide association studies (GWAS) explore the relationship between genome variability and disease susceptibility with either population- or family-based data. Here, we have evaluated the utility of combining population- and family-based statistical association tests and have proposed a method for reducing the burden of multiple testing. Unrelated singleton and parent-offspring trio cases and controls from the Genetics of Kidneys in Diabetes (GoKinD) study were analyzed for genetic association with diabetic nephropathy (DN) in type 1 diabetics (T1D). The Cochran-Armitage test for trend and the family-based association test were employed using either unrelated cases and controls or trios, respectively. In addition to combining single nucleotide polymorphism (SNP) p-values across these tests via Fisher's method, we employed a novel screening approach to rank SNPs based on conditional power for more efficient testing. Using either the population-based or family-based subset alone predictably limited resolution to detect DN SNPs. For 384,197 SNPs passing quality control (QC), none achieved strict genome-wide significance (1.4 × 10−7) using 1171 singletons (577/594 cases/controls) or 1738 pooled singletons and offspring probands (841/897). Similarly, none of the 352,004 SNPs passing QC in 567 family trios (264/303 case/control proband trios) reached genome-wide significance. Testing the top 10 SNPs ranked using aggregated conditional power resulted in two SNPs reaching genome-wide significance, rs11645147 on chromosome 16 (p = 1.74 × 10−4 < 0.05/10 = 0.005) and rs7866522 on chromosome 9 (p = 0.0033). Efficient usage of mixed designs incorporating both unrelated and family-based data may help to uncover associations otherwise difficult to detect in the presence of massive multiple testing corrections. Capitalizing on the strengths of both types while using screening approaches may be useful especially in light of large-scale, next-generation sequencing and rare variant studies.
Highlights
The successes and failures of genome-wide association studies (GWAS) have made for both interesting scientific dialog and the development of innovative statistical methodologies
The Genetics of Kidneys in Diabetes (GoKinD) study comprises 1869 type 1 diabetics (T1D) patients with and without kidney impairment who were recruited through the George Washington University Biostatistical Center (GWU) and the Joslin Diabetes Center, section of Genetics and Epidemiology (JDC)
The primary finding of this study is that analysis of GoKinD collection by any of a strict population-based design, a familybased design or the combined approach without any screening, did not detect genome-wide significant single nucleotide polymorphism (SNP)
Summary
The successes and failures of genome-wide association studies (GWAS) have made for both interesting scientific dialog and the development of innovative statistical methodologies. Many unique analytical issues have arisen with sequencing data, but two paramount themes of concern, in particular, persist regardless of the assay technology—quality control (QC) and study design We examine the latter in the context of the Genetics of Kidneys in Diabetes (GoKinD) study, a GWAS comprising one subset of unrelated subjects and another of mother-father-proband trios. Family-based studies on the other hand are robust to the discovery of spurious association due to unresolved population substructure and provide more textured information such as improved haplotype resolution, Mendelian error checking and the ability to test for imprinting effects This obviously oversimplifies the comparison of two very broad classes of designs—in this work we are concerned with implications of combining the two rather than choosing one or the other
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
