Positions, magnitudes, and distributions of amino acid residues at predictive sites selected by VIM methods.
figureposted on 2019-04-01, 17:39 authored by Craig A. Magaret, David C. Benkeser, Brian D. Williamson, Bhavesh R. Borate, Lindsay N. Carpp, Ivelin S. Georgiev, Ian Setliff, Adam S. Dingens, Noah Simon, Marco Carone, Christopher Simpkins, David Montefiori, Galit Alter, Wen-Han Yu, Michal Juraska, Paul T. Edlefsen, Shelly Karuna, Nyaradzo M. Mgodi, Srilatha Edugupanti, Peter B. Gilbert
A) IC50 censored outcome; B) Quantitative log IC50 outcome. C and D) Logo plots of the probabilities of each of the amino acids observed at key positions in (C) VRC01-sensitive Env pseudoviruses and D) VRC01-resistant Env pseudoviruses. FWER, family-wise error rate; VIM, variable importance measure. The positions illustrated here correspond to the results in Tables 1 and 2 for the presence of residues at specific sites.
CD 4 binding footprintsIC 50 neutralization titersequence featuresgp 160 pseudovirusesEnvPEAMPmethodQuantitative log IC 5026 surface-accessible residuesVRC 01 prevention efficacysieve analysisVRC 01 neutralization sensitivitygp 160 sequence featuresN-linked glycosylation sitesAUCnonparametric ensemble-based cross-validatedCATNAPdichotomous resistance outcomeHIVAntibody Mediated PreventionAAVRC 01VRC 01 neutralization