SummaryUnraveling the genetic background of common complex traits is a major goal in modern genetics. In recent years, genomewide association (GWA) studies have been conducted with large-scale data sets of genetic variants. Most of those studies have relied on single-marker approaches that identify single genetic factors individually and can be limited in considering fully the joint effects of multiple genetic factors on complex traits. Joint identification of multiple genetic factors would be more powerful and would provide better prediction on complex traits since it utilizes combined information across variants. Here we propose a multi-stage approach for GWA analysis: (1) prescreening, (2) joint identification of putative SNPs based on elastic-net variable selection, and (3) empirical replication using bootstrap samples. Our approach enables an efficient joint search for genetic associations in GWA analysis. The suggested empirical replication method can be beneficial in GWA studies because one can avoid a costly, independent replication study while eliminating false-positive associations and focusing on a smaller number of replicable variants. We applied the proposed approach to a GWA analysis, and jointly identified 129 genetic variants having an association with adult height in a Korean population.

Kawasaki disease (KD) is an acute self-limited vasculitis of infants and children that manifests as fever and signs of mucocutaneous inflammation. Coronary artery aneurysms develop in approximately 15-25% of untreated children. Although the etiology of KD is largely unknown, epidemiologic data suggest the importance of genetic factors in the susceptibility to KD. In order to identify genetic variants that influence KD susceptibility, we performed a genome-wide association study (GWAS) using Affymetrix SNP array 6.0 in 186 Korean KD patients and 600 healthy controls; 18 and 26 genomic regions with one or more sequence variants were associated with KD and KD with coronary artery lesions (CALs), respectively (p < 1 × 10(-5)). Of these, one locus on chromosome 1p31 (rs527409) was replicated in 266 children with KD and 600 normal controls (odds ratio [OR] = 2.90, 95% confidence interval [CI] = 1.85-4.54, P (combined) = 1.46 × 10(-6)); and a PELI1 locus on chromosome 2p13.3 (rs7604693) was replicated in 86 KD patients with CALs and 600 controls (OR = 2.70, 95% CI = 1.77-4.12, P (combined) = 2.00 × 10(-6)). These results implicate a locus in the 1p31 region and the PELI1 gene locus in the 2p13.3 region as susceptibility loci for KD and CALs, respectively.

Bone mineral density (BMD), the major factor determining bone strength, is closely related to osteoporotic fracture risk and is determined largely by multiple genetic factors. Semaphorin 7a (SEMA7A), a recently described member of the semaphorin family, has been shown to play a critical role in the activation of monocyte/macrophages that share progenitors with bone-resorbing osteoclasts and thus might contribute to osteoclast development. In the present study, we directly sequenced the SEMA7A gene in 24 Korean individuals, and identified 15 sequence variants.Five polymorphisms (+15667G>A, +15775C>G, +16285C>T, +19317C>T, +22331A>G) were selected and genotyped in postmenopausal Korean women (n=560) together with measurement of the areal BMD (g/cm 2 ) of the anterior-posterior lumbar spine and the non-dominant proximal femur using dual-energy X-ray absorptiometry. We found that polymorphisms of the SEMA7A gene were associated with the BMD of the lumbar spine and femoral neck. SEMA7A+15775C>G and SEMA7A+22331A>G were associated with low BMD of the femoral neck (P=0.02) and lumbar spine (P=0.04) in a recessive model. SEMA7A-ht4 also showed an association with risk of vertebral fracture (OR=1.87-1.93, P=0.02-0.03). Our results suggest that variations in SEMA7A may play a role in decreased BMD and risk of vertebral fracture.