HLA analysis and disease for multiple sclerosis

Abstract

Alleles for HLA-A, HLA-B, HLA-C, HLADQB1, HLA-DRB1, and HLA-DRB were typed across 820 age, gender, and race matched participants; 677 for whom have Multiple Sclerosis (MS) vs. 143 that do not. Observed alleles were coded and subject to significance testing with the goal to identify those alleles significantly associated with MS. Of the 368 unique alleles observed, eight were identified to be significantly associated with MS 100% of the time over 50 rounds of cross-validation testing; specifically HLA-DQB1*003, HLADQB1*00301, HLA-DQB1*00602, HLA-DRB1*0404, HLADRB1*1501, HLA-DRB1*1516/18, HLA-DRB3*02, and HLADRB4*01 were identified. Allele based classification modeling performed with each round of cross-validation testing provided for an overall classification success rate of 70%. Inspection of the alleles identified to be significantly associated with MS reveal nearly all to have been previously reported to be associated with MS; including four alleles reported to have association with disease severity. © 2011 Springer-Verlag Berlin Heidelberg.