A comparative Genome-Wide Association Interaction study using BOOST and MB-MDR algorithms on Ankylosing Spondylitis

Genome-Wide Association (GWA) studies have gained popularity after the completion of the Human Genome Project and advancement of high-throughput technologies. These studies aim to scan thousands of genomic variations (e.g., SNPs) for their association to phenotypic variables (i.e. traits), such as disease related phenotypes, with the hope of extracting biologically and clinically relevant information. Understanding of genetic, environmental as well as other components of the disease brings the key insights into disease pathology and approaches us closer to the ultimate goal - personalized medicine.
In this work we rely on a minimal GWAI protocol for genome-wide epistasis detection using SNPs, as developed in our lab [6][9]. Using the advanced non-parametric Model-Based Multifactor Dimensionality Reduction (MB-MDR) method [1] and BOolean Operation-based Screening and Testing (BOOST) algorithms [4][*] for detection of statistically significant epistatic SNP-SNP interactions, we investigate the effect of exhaustive (BOOST) and non-exhaustive (MB-MDR) marker processing strategies, LD effects, as well as different adjustment schemes for lower-order effects (i.e. epistasis).
Our approach was tested on Ankylosing Spondylitis (AS) data as provided by the WTCCC2 consortium [1]. AS is a long-term / chronic disease characterized by inflammation of the joints between the spinal bones. Non-steroidal anti-inflammatory drugs calming down the immune system inflammatory responses are used as a treatment but there is no permanent cure for AS. The disease has also a strong environmental component and affects 3.5 - 13 per 1,000 people in USA [5]