From Genome to Life:

HRABER Peter

Santa Fe Institute, 1399 Hyde Park Road, Santa Fe, New Mexico 87501, United States

title: Immunogenetics and variation in HIV progression: An application of the MDL principle

An application of computational complexity and coding theory, the minimum description length (MDL) principle states that the best hypothesis to account for some data is that which minimizes the sum of the lengths (in bits) of the description of the hypothesis and the data encoded via the hypothesis. It provides several advantages for statistical model selection, which are demonstrated with genotype data of class i and ii major histocompatibility complex (MHC) alleles and two indices of disease progression, viral load and CD4 cell counts, from a long-term HIV cohort study (Chicago MACS). Variation in viral load and CD4 cell counts is clearly associated with the MHC locus HLAB. Optimum groupings of HLAB allele supertypes associated with high, medium, and low viral load levels compare favorably with previous studies. Candidate mechanisms underlying the association include allelic variation in antigen binding specifity and frequency-dependent selection for rare HLAB alleles.