Appendix Figure 1. Phylogenetic characterization of 1300 sequences
sampled from the 9 participants described in this study and six selected
sequences from GenBank. A neighbor-joining tree was constructed using
maximum-likelihood distances in Phylip v3.5 (1)
after regions that could not be unambiguously aligned were removed. Sequences
sampled from each study participant formed monophyletic clusters separate from
the other participants as well as from the six unrelated sequences included in
this illustration. No viral sequence from this study was found to be closely
related to prototypic HIV-1 sequences or to other known strains present or
sequences determined in the laboratory. These observations are consistent with
an absence of sample contamination or mix-up (3).
 |
 |
 |
 |
 |
 |
Appendix Figure 2. Pairwise comparisons of virus population divergence
from the founder population (panels on the left side of the figure) and
diversity within each sample (panels on the right) in each of the nine
participants. Pairwise distances were estimated using the Kimura 2-parameter
model of viral evolution (2).
These distances were plotted on the y-axis at the corresponding times following
seroconversion. The lines within each panel connect the mean values for each
time point. Comparisons of sequences from PBMC are shown with a red open
circle for each data point, those from plasma are shown with a blue vertical
line for each data point. The first 5 panels show the estimates of viral
diversification for participants in which sequences were sampled from both PBMC
and plasma. Subsequent panels show these estimates sequences sampled from PBMC
DNA or from plasma RNA (but not both at any time point). You may click on any
panel in the figure above to see an enlarged version or reach a link so that you
can download it in PDF format.
 |
 |
 |
Appendix Figure 3. Phylogenetic analysis of HIV-1 env C2-V5
sequences from each of the nine participants. Sequences were aligned and
visually adjusted using GDE (4).
The phylograms were estimated by the method of neighbor-joining from a matrix of
pairwise maximum likelihood sequence distances (transition/transversion ratio =
1.4) using programs from version 3.5 of the PHYLIP package (1).
Trees were rooted to one of the sequences sampled at the first time point using
Treeview (from R.
Page). The legend for each phylogram shows the time of sampling in months
following seroconversion and is depicted in an arbitrary color gradient. DNA
sequences from PBMC are depicted as squares and plasma RNA sequences as
triangles. Sequence changes expected to confer an X4 phenotype (SI phenotype on
MT-2 cells and CXCR4 co-receptor specificity) based on basic amino acid
substitution at any of the three positions (#306, 319, 320 of gp120) are
indicated by a red asterisk. You may click on any panel in the figure above to
see an enlarged version or reach a link so that you can download it in PDF
format.
Appendix Figure 4. Comparison of DNA distance estimations in nine
participants using Kimura 2-parameter (K2P) model of evolution with a transition
to transversion ratio of 2 (2)
vs. a General Time Reversible (GTR) model with a site-to-site variation in
substitution rates (discrete approximation of a 
distribution with a
shape parameter, 
=0.5) (5).
DNA distance estimates are shown for sequences from PBMC only in all
participants except for Participant1, where plasma data is shown for the time
points when PBMC samples were not available, and Participant 11 for whom
insufficient amounts of viral DNA was found in PBMC. K2P distances are plotted
as blue circles and GTR DNA distances are plotted as red triangles. Intra-time
point DNA diversity is depicted by open symbols and the DNA distance as compared
to founder sequences is shown by filled symbols. These comparisons show similar
trends using either K2P or the GTR distances, and as expected, at high levels of
DNA differences, GTR model yields a comparatively higher DNA distances.You may
click on the figure above to see an enlarged version or reach a link so that you
can download it in PDF format.