HIV Vaccine Research
David Heckerman recognized that email SPAM and the HIV virus share common traits – both mutate constantly, have many potential variations and spread quickly. HIV researchers faced constraint as it is difficult to analyze the large virus mutation data sets. The 8,121,791 patent covers an algorithm known as “PhyloD” that sequences each HIV genome and catalogues potential weak links among nucleotides in each sample for comparison. Specifically, the patent describes a trained logistic regression model, e.g., exemplary classification model, to execute logistic functions on received protein data. The model also incorporates hidden binary variables and shift variables (used for MHC type-II allele prediction) that, when processed, represent the identification, e.g, prediction, of desired epitopes. The classification model can be configured to predict epitope information by processing data including various features of an epitope, Major Histocompatibility Complex (MHC), MHC supertype, and Boolean combinations.