Features of Recently Transmitted HIV-1 Clade C Viruses that Impact Antibody Recognition: Implications for Active and Passive Immunization
Fig 2
K-means (n = 4) neutralization clustering of clade C Env pseudoviruses for tier classification (n = 200).
Neutralization sensitivity was assessed by assaying pseudoviruses against 30 sera from chronically infected individuals. Individual viruses are indicated to the right of the heatmap with individual serum listed at the bottom. Viruses are ranked according to respective sensitivities and categorized into tiers as determined by k-means clustering. Neutralization sensitivity (log10 ID50 titer) is denoted by color key provided in the inset. Viral isolates clustering, together with >95% probability, was performed to assess the stability of the Env pseudovirus clusters based on a random-with replacement resampling of the sera (rows), or to assess the stability of the serological clusters based on resampling of the viruses (columns), and were boxed on the heatmap indicating categorization into their respective tiers. Bootstrap support for clusters was determined by resampling the data sets from the individual serum and re-evaluating the k-means clusters 10,000 times giving an indication of how many times each virus was a member of the originally assigned tier classification in the resampled datasets. Env pseudoviruses that clustered with the assigned tier are shown in the column on the left as blue for tier 1A (very sensitive), yellow for tier 1B, magenta for tier 2 (intermediate), and black for tier 3 (very resistant). This follows the tier classification of Seaman et al (2010) [16]. Similarly sera potential is illustrated at the top with the most uniformly potent sera indicated as blue, potent sera as yellow, modestly potent as magenta, and least potent as black. Color intensity indicates the probability of falling within a cluster and the degree of blending the frequency of falling into each of the respective tiers, with strong unblended colors, associated with 95% bootstrap support, boxed. The poor resolution of the k- means bootstrapping emphasizes that the sensitivity of Envs is a continuum, and will be dependent on the specific sera or antibodies used for the evaluation.