Search

We hypothesize that patterns of sexual behavior play a role in the conformation of transmission networks, i.e., the way you behave might influence whom you have sex with. If that was the case, behavioral grouping might in turn correlate with, and potentially predict transmission networking, e.g., proximity in a viral phylogeny. We rigorously present an intuitive approach to address this hypothesis by quantifying mapped interactions between groups defined by similarities in sexual behavior along a virus phylogeny while discussing power and sample size considerations. Data from the Swiss HIV Cohort Study on condom use and hepatitis C virus (HCV) sequences served as proof-of-concept. In this case, a strict inclusion criteria contrasting with low HCV prevalence hindered our possibilities to identify significant relationships. This manuscript serves as guide for studies aimed at characterizing interactions between behavioral patterns and transmission networks. Large transmission networks such as those of HIV or COVID-19 are prime candidates for applying this methodological approach.

Little is known about whether and how variation in the HIV-1 genome affects its transmissibility. Assessing which genomic features of HIV-1 are under positive or negative selection during transmission is challenging, because very few virus particles are typically transmitted, and random genetic drift can dilute genetic signals in the recipient virus population. We analyzed 30 transmitter-recipient pairs from the Zurich Primary HIV Infection Study and the Swiss HIV Cohort Study using near full-length HIV-1 genomes. We developed a new statistical test to detect selection during transmission, called Selection Test in Transmission (SeTesT), based on comparing the transmitter and recipient virus population and accounting for the transmission bottleneck. We performed extensive simulations and found that sensitivity of detecting selection during transmission is limited by the strong population bottleneck of few transmitted virions. When pooling individual test results across patients, we found two candidate HIV-1 genomic features for affecting transmission, namely amino acid positions 3 and 18 of Vpu, which were significant before but not after correction for multiple testing. In summary, SeTesT provides a general framework for detecting selection based on genomic sequencing data of transmitted viruses. Our study shows that a higher number of transmitter-recipient pairs is required to improve sensitivity of detecting selection.