Sequence analysis can be used to evaluate transmission networks. We have used retrospective samples to examine two HIV-1 transmission networks established by contact tracing. Regions of the HIV-1 region representing segments of gag and env were amplified by RT-PCR from frozen plasma samples and the sequence of each PCR product was determined. Within one of the networks (composed of 38 subjects) we found only a subset of the tested sequence clusters was consistent with the reported

more »

... l linkage. Of 15 presumed transmission events where sequence data were available, 9 could be rejected either by subtype mismatch or by phylogenetic tests. In the other network (composed of 89 subjects) we were able to assess sequences for 26 presumed transmission events, 18 of which were rejected based on subtype discordance. Long lags in time between the time of transmission and the time of sequence sampling (ranging from 2 to 18 years) may limit the sensitivity for the detection of sequence linkage. Also, superinfection and incomplete epidemiological information are other factors that will limit the concordance of phylogenetic reconstruction and reported epidemiological linkage.