Estimating the respective contributions of human and viral genetic variation to HIV control. PLoS Computational Biology
Bartha et al. evaluated the fraction of variation in HIV-1 set point viral load (spVL) attributable to viral or human genetic factors by using joint host/pathogen genetic data from 541 HIV infected individuals. On the host side, the authors focused on amino acid variations in the HLA-A, B and C genes. They built three linear mixed models, one containing human variants, one derived from phylogenetic trees, and one including both host and virus information.
The genetic relatedness matrix created from 33 amino acid polymorphisms of the human class I HLA genes explained 8.4% (standard deviation = 4%) of the observed variance in spVL. In contrast, 28.8% (SD = 11%) of phenotypic variation was attributable to the viral phylogenetic tree. Combining the two relatedness matrices in one model yielded a total variance explained of 29.9% (SD = 12%) less than the sum of the latter two models.
In conclusion, the study shows that a major fraction of inter individual variability is explained by the similarity of the viral genotypes, and that human genetic variation in the HLA region provide little additional explanatory power. The results suggest that host genetic association studies not taking the virus into account underestimate the population level effect of host genetic variation. Combining host and pathogen data provides additional insight into the genetic determinants of the clinical outcome of HIV infection, which can serve as a model for other chronic infectious diseases.