An R package for quality control of genome-wide association study results

Genome-wide association studies (GWAS) have significantly changed our view on the genetic architecture of human phenotypes and complex-trait genetics within the broad scope of systems biology. This approach provides an unbiased, hypothesis-free method to discover genetic variants across the human genome that are associated with a disorder or phenotype. Given that millions of markers are tested for association in a GWA study, the smallest flaw or systematic bias in the data can lead to misleading findings. This highlights the role of data preparation and quality control (QC) as an essential step of such analysis. The objective of QC is to ensure that results files are valid, complete, correctly formatted, and, in case of meta-analysis, consistent with other studies in the same analysis. Due to complexity of the data, numerous QC metrics calculations, sheer size of data files and the risk of analyst bias, this step is not feasible by hand and should be automated to save time and prevent human errors.Here, we present an R package that was developed to facilitate and pipeline this process. GWASinspector is a cross-platform package with minor external dependencies that employs object oriented programming via S۴ object models in R programming language and is publicly available from the Comprehensive R Archive Network (CRAN) (۱). The ability to process insertion/deletion and multi-allelic variants, handling big data in an efficient manner, and comprehensive graphic reports are the main strengths of this software compared to the existing packages (۲–۵). Required databases, including population-specific allele frequency reference datasets and variant effect-size reference datasets, and a detailed tutorial are available on our website at http://GWASinspector.com.

کلیدواژه ها:

Systems Biology ، Genome-Wide Association Study ، Quality Control ، Computer Software ، R