Systemic lupus erythematosus (SLE) is characterized as an autoimmune disorder with unclear etiology. To identify the genetic effect of SLE, a genome wide association study (GWAS) and its further replication were conducted on SLE patients in Asian populations and ethnically matched controls. Before this study, most of the confirmed association loci were identified by GWAS studies in European populations. Apart from the established associations, we identified a SLE susceptible single nucleotide polymorphisms (SNP) located in ETS1 (rs1128334, P=2.3E-11, OR=1.29) in four different cohorts. This locus is probably an Asian-specific susceptibility locus since no Caucasian study has reported further validation in the last years. A new susceptibility variant in UHRF1BP1 (rs13205210, P=4.4E-09, OR=1.49) independent from the previously confirmed SNPs in Caucasian study was also confirmed to be associated with SLE in the Hong Kong Chinese population. Meta-analysis was performed by introducing another Chinese Han GWAS data set from Anhui province, China. Three loci, TET3-DGUOK, CD80, DRAM1, were confirmed to be associated with SLE. Two loci with suggestive signals in Hong Kong GWAS and further replication were also confirmed by the meta-analysis: PTTG1-MiRNA146a, YDJC. In order to identify the genetic effect for females who have predominant chance to suffer from SLE, X chromosome specific meta-analysis based on the Hong Kong and Anhui GWAS data and further replication study were performed by considering the difference between females and males. A signal in PRPS2, and three independent signals in the Xq28 were confirmed with the replication in three different cohorts by considering both females and males. Gene-gene interactions were also investigated genome-widely in a hypothesis free manner based on the meta-analysis results. The further validation processes were preceeded based on each independent GWAS data set. Four pairswise interacting loci were found and cross validated by three methods including logistic regression and Multifactor Dimensionality Reduction (MDR) and information gain theory based on the Anhui GWAS data set. Further studies are still needed to better explain the real features of genetic epistasis and the potential biological roles. By incorporating two GWAS from the same population, the population difference is efficiently avoided. Together with the putative gene-gene interactions, this study presents a comprehensive analysis based on the GWAS data conducted on SLE. It may shed new light on the disease mechanisms of SLE. / published_or_final_version / Paediatrics and Adolescent Medicine / Doctoral / Doctor of Philosophy

published_or_final_version / Paediatrics and Adolescent Medicine / Master / Master of Research in Medicine

SLE is a complicated autoimmune disease with a strong genetic basis. Owing to the successful application of genome-wide association study (GWAS) on complex traits, significant progresses have been made in identifying susceptible loci for SLE. So far, more than 30 loci have been shown to have robust association with the disease, significantly improving our understanding of this disease. On the other hand, these loci only confer relatively small increments in disease risk, leaving a large amount of disease heritability unexplained. Several limitations may lead to the issue of “missing heritability”, such as the insufficient statistical power of standard GWAS, the failure to explore independent signals with marginal effects, and other forms of genetic variations or interactions that can hardly be detected by the current GWAS approach. In the current study, efforts have been made on solving the issue of missing heritability. Firstly, in order to increase the statistical power of GWAS, we performed meta-analysis of two existing SLE GWASs on Chinese populations, and replicated the findings in additional samples from Hong Kong, Anhui, and Thailand. We identified ARID5B and CDKN1B as novel SLE susceptibility genes through this approach. Secondly, considering the limitation of previous GWASs that only focus on the most significant signal in an individual region, we revisited the established loci by in-depth analysis on the basis of meta-analysis and followed up the findings with large-scale replication efforts. We found three additional independent signals in 11q23.3 as being associated with SLE. Thirdly, given the importance of independent effects in predisposing disease susceptibility, we performed gene-based analysis to combine the marginal independent signals within a gene to identify novel susceptibility genes. Our results demonstrated the widespread existence of independent effects within known SLE susceptibility genes, and we also identified ANXA6 as a novel SLE susceptibility gene. Lastly, we also made some efforts in the characterization of the related genes, and we identified several regulatory SNPs (rSNPs) predisposing individuals to SLE via affecting expression of the corresponding genes. In addition, we also found that epistatic interaction of variants in a previously established susceptibility gene, ETS1,correlates with cytokine (e.g.IL-17) abnormality in SLE cases. The current work represents several practical approaches to improve the power of GWAS. The findings might enrich the list of SLE susceptibility genes as well as advance our knowledge on the etiology of this complicated disease. Importantly, the current study highlights the importance of independent effects and rSNPs in conferring disease susceptibility, which may shed light on further exploration on the genetic architecture of SLE. / published_or_final_version / Paediatrics and Adolescent Medicine / Doctoral / Doctor of Philosophy

published_or_final_version / abstract / toc / Paediatrics and Adolescent Medicine / Master / Master of Philosophy

published_or_final_version / abstract / Paediatrics and Adolescent Medicine / Master / Master of Philosophy

published_or_final_version / abstract / Paediatrics and Adolescent Medicine / Master / Master of Philosophy