Table 3.
Forward stepwise regression models for anti-La
Marker | Estimates from the multiple regression model
|
Single marker
|
|||||
---|---|---|---|---|---|---|---|
OR | 95% CI | P-value | Class | Gene | OR | P-value | |
(A) Forward stepwise regression (BIC = 1677.92) | |||||||
rs2894254 (C<A) | 3.38 | 2.74–4.16 | 3.40 × 10−30 | Class III | c6orf10—BTNL2 | 3.38 | 3.40 × 10 −30 |
(B) Conditional on HLA-DRB1*03:01 as a covariate (BIC = 1673.31) | |||||||
HLA-DRB1*03:01 | 2.50 | 2.00–3.13 | 1.40 × 10−15 | HLA-DRB1 | 3.15 | 3.31 × 10 −28 | |
rs9268832 (C<A) | 1.64 | 1.32–2.04 | 6.53 × 10−06 | Class II | DRA-DRB5 | 2.31 | 2.46 × 10 −17 |
(C) Forward stepwise regression starting from AA-DRB1*77T (BIC =1672.53) | |||||||
Thr77 DRB1 | 0.40 | 0.32–0.50 | 1.33 × 10−15 | HLA-DRB1 | 0.32 | 2.40 × 10 −28 | |
rs2227139 (C<A) | 1.64 | 1.32–2.04 | 6.47 × 10−06 | Class II | HLA-DRA | 2.32 | 1.86 × 10 −17 |
Abbreviations: BIC, bayesian information criterion; CI, confidence interval; OR, odds ratio. SNPs have their minor and major alleles noted in brackets (A<G where A is the minor allele, for example), the OR is with respect to the minor allele. For BIC see Materials and Methods.