Return to search

An Analysis of Genome-Wide Association Studies to Produce Evidence Useful in Guiding Their Reporting and Synthesis

Introduction The present study evaluated reported methodological characteristics of GWAS, investigating relationships between reported methodological characteristics and outcomes observed.
Methods GWAS were identified from NHGRI’s catalogue of GWAS (2005 to 2009). Multivariate meta-regression models (random effects) were produced to identify the impact of reported study characteristics and the strength of relationships between the variables and outcomes.
Results The summary odds ratios for replication components of GWAS in cancer was 1.34 (95% CI 1.25, 1.43) and neuropsychiatric disorders was 1.43 (95% CI 1.30, 1.57). Heterogeneity was accounted for by nature of the control group, relationship between case/control groups, whether cases/controls were drawn from the same population, if data was a primary collection or a build on pre-existing data, if quality assurance was reported, and if the study reported power/sample size.
Conclusion Evidence supports the existence of variability in reporting, with index components demonstrating less variability than replication components in the GWAS.

Identiferoai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:OOU-OLD./20686
Date08 February 2012
CreatorsYurkiewich, Alexander John
Source SetsLibrary and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada
LanguageEnglish
Detected LanguageEnglish

Page generated in 0.0016 seconds