The ability of HIV to evade antibodies through various mutational pathways poses a major hurdle in the development of an effective vaccine. An ideal solution involves broadly neutralizing antibodies that target sensitive virus spike proteins, in which any arising mutations would severely compromise the virus’s fitness. To achieve this, a team led by researchers from Hong Kong University of Science and Technology and MIT have developed a computational approach that maps the fitness landscape of one such spike, the HIV gp160 polyprotein.
The team processed data comprising 815 amino acid residues and 20,043 protein sequences from 1,918 HIV-infected individuals. Using a computational method, the team statistically inferred the fitness landscape—defined as fitness as a function of sequence—of the polyprotein. Through comparisons with diverse experimental measurements, the team then validated this inferred landscape.
"Without big data machine learning methods, it is simply impossible to make such a prediction," said co-author Raymond Louie. "The number of parameters needed to be estimated came close to 4.4 million."
The team’s paper was published recently in the Proceedings of the National Academy of Sciences (PNAS). It concluded that the availability of this fitness landscape model can aid in the rational design of immunogens for effective vaccines.
"The computational method gave us fast and accurate results," said co-author Matthew McKay. "The findings can assist biologists in proposing new immunogens and vaccination protocols that seek to force the virus to mutate to unfit states in order to evade immune responses, which is likely to thwart or limit viral infection."
“Our fitness landscape could be clinically useful in the future for the selection of combination bnAb therapy and immunogen design," said co-author Arup Chakraborty.
Image: Schematic of the computational method for estimating the fitness landscape of HIV-envelope. Image courtesy of Department of Electronic & Computer Engineering, HKUST.