ABSTRACT Systemic lupus erythematosus (SLE) is a heritable autoimmune disease that primarily affects young females. SLE symptoms can be very heterogeneous, which posts great challenges in early diagnosis. In its advanced stages, SLE can lead to multiple organ failures and even fatality. Early diagnosis is critical for controlling and mitigating the symptoms and improving the quality of life. Genome-wide association studies of SLE to date have identified >150 loci. Yet, the causal variants remain elusive for most loci. There is great interest to integrate datasets from diverse human populations to further empower discovery, refine causal variants, and improve the risk prediction accuracy. In this proposal, we seek to aggregate available GWAS summary statistics from a myriad of autoimmune diseases. We will develop improved meta-analysis methods that effectively integrate data from multiple traits and ancestries. We will also develop better fine-mapping methods that can integrate statistical approaches and experimental validations. Finally, we will develop more accurate genetic risk scores using datasets from multiple ancestries and traits, and combine them with commonly used lab values for improved disease risk predictions. We will release useful software packages implementing these methods to benefit studies of other traits and maximize the impact of the proposed research.