An amino acid motif in HLA-DRß1 distinguishes patients with uveitis in juvenile idiopathic arthritis

A.J.W. Haasnoot; M.W. Schilham; S.S.M. Kamphuis; P.C.E. Hissink Muller; A. Heiligenhaus; D. Foell; R.A. Ophoff; T.R.D.J. Radstake; A.I. Den Hollander; T.H.C.M. Reinards; S. Hiddingh; N. Schalij-Delfos; E.P.A.H. Hoppenreijs; M.A.J. van Rossum; C. Wouters; R.K. Saurenmann; N. Wulffraat; R. ten Cate; J.H. de Boer; S.L. Pulit; J.J.W. Kuiper

omnibus test of all but one of the 6 alleles present at HLA-DRB1 position 11 compared to a null model. We found that the goodness-of-fit of the omnibus test far exceed that of a null model including only sex, principal components and genotyping batch (likelihood ratio test p = 1.5 x 10 -9 ). To test if a single HLA-DRB1 position 11 allele was driving the top association signal (as opposed to all possible alleles together, as modeled by the omnibus test), we conditioned the top association signal first on serine and then on aspartic acid at position 11, by including the dosages of the respective variant as an additional covariate in the logistic regression model (Figure 1 , Table 2, Supplementary Table 2 Table 2) . To further decipher the specific residue(s) driving the association in uveitis, we performed a series of likelihood ratio tests. In brief, a likelihood ratio test compares the fit of two models to the observed data and indicates which of the models, if either, best fits the data. We found that the S (or Y or T) amino acids best fit the data (p = 1.48 x 10 -10 , compared to a model containing only PCs and sex as independent variables). Including the S/G residue at position 13 as well as the YST motif only modestly improved the fit of the model (likelihood ratio test p = 0.043, compared to a model containing serine at position 11 only); neither the S nor G residues alone at position 13 improved the model (p = 0.70 for both tests).