|
Abstract:
|
We develop an antigen scanning method in order to evaluate the results from HIV vaccine efficacy trials. An HIV amino acid sequence is measured from each volunteer who acquires HIV during the trial. These sequences are multiply aligned, along with the prototype HIV sequence represented in the vaccine, to form a high-dimensional genomic dataset. Based on these data, we identify short, 8-12 amino acid long, peptide regions at which the peptide sequences from infected vaccinees tends to be more divergent from the prototype vaccine sequence than the sequences from infected placebo recipients. We approach this problem by extending earlier methods for comparing positions, using Euclidean and Kullback-Leibler-type metrics for measuring peptide distances. The methods are illustrated by a simulation study and with data from an HIV vaccine efficacy trial.
|
