There are more than 80 different autoimmune diseases, and collectively they affect 3-5% of the population. They are chronic, incurable, and cause substantial morbidity and mortality. The fundamental lack of understanding of autoimmune disease pathogenesis means that available treatments – most causing general immune suppression - are non-specific, often fail to treat the most serious disease manifestations, and cause serious side-effects. Our research over the past decade first focused on “mouse-to-human” approaches to understand autoimmunity. Random mouse mutagenesis led to our discovery of the Roquin family revealing a pathway in which accumulation of Tfh cells caused by Roquin mutations leads to rogue positive B cell selection and autoantibody-mediated autoimmune disease. Roquin and its paralogue Roquin2 act in T cells and myeloid cells to repress mRNAs posttranscritptionally. The key pathogenic pathways leading to autoimmunity and inflammation when Roquin is mutated consist on excessive IFN-g production that promote Tfh and GC B cell accumulation, and increased myeloid cell-derived TNF. More recently, we have undertaken a “human-mouse-human” approach to focus our efforts on illuminating the mechanism of disease in each individual patient and elucidating the biological pathways that connect human genetic variation to disease. This has led to identification of rare gene variants contributing strong effects to autoimmune susceptibility. Amongst these is a case of childhood cerebral SLE manifesting in the first years of life with severe clinical manifestations including stroke. A novel homozygous variant in TREX1 leading to a significant reduction in the protein’s exonuclease activity causes increased type I interferon production, likely to be at the root of the autoimmune phenotype. This human-centered approach is expected to help stratify what are notoriously heterogeneous diseases, and illuminate targeted therapies based on the affected molecular pathways.