Methods study for number of controls per case to choose in GWAS
Principal Investigator
Name
Hormuzd Katki
Degrees
PhD
Institution
NCI
Position Title
Senior Investigator
Email
About this CDAS Project
Study
PLCO
(Learn more about this study)
Project ID
PLCO-1200
Initial CDAS Request Approval
Apr 13, 2023
Title
Methods study for number of controls per case to choose in GWAS
Summary
The rule of thumb that there is little gain in statistical power by obtaining more than 4 controls per case, is based on type-1 error alpha=0.05. We have shown, when alpha is very small, such as genomewide significance, there is substantial gain in statistical power by obtaining more than 4 controls per case, such as 10 or more. We request phenotypes for 3 SNPs, 2 associated with melanoma (rs6059655, rs871024) and one with pancreas cancer (rs635634). We will examine how the OR varies as the number of controls per case increases. We have submitted to DUPS, #68.
Aims
We have submitted to DUPS, #68. We want to use the PLCO GWAS Explorer data for 3 SNPs, 2 associated with melanoma and 1 with pancreas:
All Melanoma (European) - 8,951,271 variants, 2,629 cases, 67,500 controls
Chr. Pos. SNP
20 34077942 rs6059655
9 21803881 rs871024
Pancreatic Cancer (European) - 8,954,671 variants, 674 cases, 67,500 controls
Chr. Pos. SNP
9 133279427 rs635634
We would like the cancer phenotype with the GWAS Explorer SNP data and examine how increasing the number of controls per case affects the OR and its p-value. Jianxin Shi and Wendy Wong (NCI) are leading the analysis.
Collaborators
Jianxin Shi (NCI)
Wendy Wong (NCI)