Fig. 4

Genotypic signature for IUT and BMT groups. a Comparison of gp160 length (in number of amino acids) and b glycosylation level (in number of PNGs) for maternal and infant envelopes, segregated by route of transmission. The number of envelope sequences tested for each group (n) is given. c–d Comparison of infant and maternal envelope sequence length and PNG content, stratified by transmission mode and by gp120 variable domains (V), constant domains (C), and the gp41 subunit. Each bar along the horizontal axis represents an individual transmission pair, which is solid if comparison between the individual mother–infant pair is significant (p < 0.05) and outlined if not significant*. The y-axis represents either the difference in sequence length or the difference in number of PNG sites found between infant and maternal envelope clones, computed as [mean maternal isolate length (or PNG content)—the mean infant isolate length (or PNG content)] for each pair. p values were calculated using GEE for each individual domain, and domains that differed significantly for corresponding maternal and infant isolates are boxed in red. The graph in c shows differences in sequence length and PNG level for in utero transmission pairs, and d shows differences in length and PNG level for breast milk transmission pairs. *Wilcoxon rank-sum tests were used to compare sequence length or PNG content found in maternal and infant isolates from a given transmission pair, with resulting p values being corrected for multiple comparisons using the method of Benjamini and Hochberg [105]