Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2013 Dec 2;8(12):e80562.
doi: 10.1371/journal.pone.0080562. eCollection 2013.

Computational prediction of broadly neutralizing HIV-1 antibody epitopes from neutralization activity data

Andrew L Ferguson  1 , Emilia Falkowska, Laura M Walker, Michael S Seaman, Dennis R Burton, Arup K Chakraborty
Affiliations

Computational prediction of broadly neutralizing HIV-1 antibody epitopes from neutralization activity data

Andrew L Ferguson et al. PLoS One. .

Abstract

Broadly neutralizing monoclonal antibodies effective against the majority of circulating isolates of HIV-1 have been isolated from a small number of infected individuals. Definition of the conformational epitopes on the HIV spike to which these antibodies bind is of great value in defining targets for vaccine and drug design. Drawing on techniques from compressed sensing and information theory, we developed a computational methodology to predict key residues constituting the conformational epitopes on the viral spike from cross-clade neutralization activity data. Our approach does not require the availability of structural information for either the antibody or antigen. Predictions of the conformational epitopes of ten broadly neutralizing HIV-1 antibodies are shown to be in good agreement with new and existing experimental data. Our findings suggest that our approach offers a means to accelerate epitope identification for diverse pathogenic antigens.

PubMed Disclaimer

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1
Figure 1. Logo plot of the variability of selected positions in HIV-1 Env within the 141-strain pseudovirus panel.
We present data for all positions identified in Tables 1–3 as significant determinants of bnMAb neutralization activity by either the ensemble classifier or experimental alanine scan data.
Figure 2
Figure 2. Compressed sensing (CS) selection of PGT-123 epitope residues.
Results of the application of the compressed sensing classification algorithm to the neutralization activity of bnMAb PGT-123 against a panel of 141 HIV-1 pseudoviruses (cf. Table S1). In each panel, the abscissa indicates the number of non-zero elements in the formula image signal vector computed by the LASSO algorithm, and therefore the number of residues incorporated into the regularized least squares fit of the neutralization data (Eqn. 3). For clarity of viewing, plots are terminated at the 100-component model. As indicated by the arrows, knees in the (a) mean squared error (MSE) over the complete data set and (b) leave-one-out cross-validation mean squared error (LOOCV-MSE) curves were identified using the L method at 11 and 9 residues, respectively . The mean of these values motivated the selection of the ten residues constituting this model: I323, H330, N332, N334, S334, S612, N671, Q740, V815, and V843 (c.f. Table 1).
Figure 3
Figure 3. Mutual information (MI) selection of PGT-123 epitope positions.
The redundancy spectrum produced by application of the mutual information classification algorithm to the neutralization activity of bnMAb PGT-123 against a panel of 141 HIV-1 pseudoviruses (cf. Table S1) using an IC50 cutoff of 10 µg/ml. The ordinate records the computed redundancy of the residue identity in each position with the observed neutralization activity. The abscissa lists the positions of the protein in decreasing order of redundancy. The dashed line indicates the cutoff computed by the shuffling procedure described in Materials and Methods, Rcutoff = 0.15, above which redundancy values should be considered statistically significant. These results suggest that the three top ranked positions – respectively, 332, 334 and 330– be retained in the model (cf. Table 1). For clarity of viewing, plots are terminated at the 100-component model.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. De Cock KM, Jaffe HW, Curran JW (2011) Reflections on 30 years of AIDS. Emerging Infectious Diseases 17: 1044–1048. - PMC - PubMed
    1. Rerks-Ngarm S, Pitisuttithum P, Nitayaphan S, Kaewkungwal J, Chiu J, et al. (2009) Vaccination with ALVAC and AIDSVAX to prevent HIV-1 infection in Thailand. New England Journal of Medicine 361: 2209–2220. - PubMed
    1. Plotkin SA (2010) Correlates of protection induced by vaccination. Clinical and Vaccine Immunology 17: 1055–1065. - PMC - PubMed
    1. Plotkin SA (2008) Correlates of vaccine-induced immunity. Clinical infectious diseases 47: 401–409. - PubMed
    1. Amanna IJ, Messaoudi I, Slifka MK (2008) Protective immunity following vaccination: how is it defined? Human Vaccines 4: 316–319. - PMC - PubMed

Publication types

Grants and funding

Financial support for this work was provided by the Ragon Institute (A.K.C.; http://www.ragoninstitute.org), a National Institutes of Health Director’s Pioneers Award (A.K.C.; https://commonfund.nih.gov/pioneer), and a Ragon Postdoctoral Fellowship (A.L.F.; http://www.ragoninstitute.org). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.