Fig 1.
Association of SNPs (black), amino-acids (AA; red), and classical alleles (blue). b. Accumulation of risk residues.
Odds ratios estimated as a comparison between the number of risk residues in cases and controls in discovery and replication cohorts.
Table 1.
Independent HLA association. P-values from Log-likelihood ratio tests of the combined model (conditioned on three principal components, sex and ethnic background).
KR: Korean; HC1/2: Han Chinese 1/2 (1 has added controls); MC: Malaysian Chinese; JP1/2: Japanese cohorts 1/2. Fisher’s combined P-value is presented to combine individual cohorts’ omnibus-association values. Residues are presented in the order they entered the conditional regression model. Values for DRB1-11 and DRB1-13 are unconditioned.
Fig 2.
Accumulation of risk residues.
Odds ratios estimated as a comparison between the number of risk residues in cases and controls in discovery and replication cohorts.
Table 2.
Significant mapped SLE risk residues.
The most-risk (highest OR; in 19/20 cases, also the most significant) side-chain at each protein position is evaluated according to a simple model (Supplementary Note, S10a Table). Side-chains predicted risk include: W, I, F, L, E, Y, D. R, S, N, K, T are predicted to be protective. Residues written as: position, amino acid, position in binding pocket (p1-p9 or peptide-binding 2nd shell), P-value for the discovery cohort, odds ratio. *: DRB1-13R is in complete LD with DRB1-11P. Residues are formatted according to their consistency with the statistical SLE risk-prediction model: risk (bold); protective (italics); and neither (unformatted). Amino acid properties: Negatively charged (Asp, D; Glu, E); Positively charged (Arg, R; Lys, K); Potentially positively charged (His, H); Large, hydrophobic (Trp, W; Ile, I; Phe, F; Leu, L; Tyr, Y; Met, M); Medium, hydrophobic (Pro, P; Val, V); Small, hydrophobic (Ala, A; Gly, G; Cys, C); Neutral hydrophilic (Gln, Q; Asn, N; Ser, S; Thr, T).
Table 3.
DRB1 risk and protective residues for specific autoantibodies.
Significant positions map overwhelmingly to Class II proteins, primarily DRB1 for nRNP, Ro/La and cardiolipin, and DPB1 for Sm. Positions with significant association have the most-risk and most-protective amino acids shown. Residues written as: position, amino acid, position in binding pocket, P-value for the Korean and Han Chinese samples, odds ratio. Side-chains predicted risk include: F, L, D. R and S are predicted to be protective. Residues are formatted according to their consistency with the statistical SLE risk-prediction model: risk (bold); protective (italics); and neither (unformatted).
Table 4.
OR: Odds-ratio; LCI: Lower confidence limit; UCI: Upper confidence limit.
Table 5.
Significantly associated residues across eleven autoimmune diseases within the peptide-binding groove of HLA-DRB1.
Bold red amino-acids denote independently associated risk signals in the strongest associated DRB1 allele, identified by the authors of each study. Bold green amino-acids denote significant protective signals. SLE: lupus, MS: multiple sclerosis, RA: rheumatoid arthritis, T1D: type 1 diabetes, HT: Hashimoto’s thyroiditis, pSS: primary Sjögren’s syndrome, SSc: systemic sclerosis, UC: ulcerative colitis, CD: Crohn’s disease, PV: pemphigus vulgaris, Vi: vitiligo. *: Complete linkage disequilibrium (D’ = 1).
Table 6.
Rheumatoid arthritis and systemic lupus erythematosus immunogenic aminoacids.