Quantitative Delineation of Herpesviruses in Bats for use in Ecological Studies

Anna R. Sjodin; Michael R. Willig; Simon J. Anthony

where i and j represent index values of host species, p represents OTU prevalence within host populations, and w ij is the phylogenetic distinctiveness of hosts. To measure phylogenetic distinctness, hosts of the same genus are assigned a value of one, hosts of the same family but different genera are assigned a value of two, and hosts of different families have a value of three (Poulin and Mouillot 2005). S TD * thus weighs host preference by the relatedness of hosts (incorporating evolution) and the prevalence of the pathogen in each of its hosts (incorporating ecology). The extent of viral sharing among hosts species was further examined by determining whether particular OTUs had a core host species. A core host species was defined as a single host species for which the OTU had significantly higher prevalence than in all other host species . CC-BY-NC-ND 4.0 International license is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It . https://doi.org/10.1101/856518 doi: bioRxiv preprint combined. To do this, the prevalence of each OTU in its core host and its combined prevalence in all other hosts were compared using Fisher's exact test (Wassertheil-Smoller and Smoller 2015) . For each OTU that infected more than one host individual, viral preference for host sex was examined by comparing OTU prevalence between male and females of the core host species, using Fisher's exact test. All analyses were done using R version 3.4.3, and code used for all methods can be found at github.com/ARSjodin. OTU Delineation: Both alignment techniques resulted in two similar, clearly differentiated potential cutoff ranges, as demonstrated by consistent peaks in the histograms (Figure 1 ). The lower level of taxonomic grouping of OTUs was defined as those with a cutoff between 90 -92%. This means that two sequences with less than 90% genetic similarity would be considered to be different OTUs. The higher level of taxonomic grouping was defined as those with a cutoff between 74 and 80%. As such, two sequences with less than 74% genetic similarity would be considered to be different OTUs. The number of OTUs resulting from the cutoff extremes for both levels of classification, when total alignment and pairwise alignment were considered (74, 81, 90 and 93 PID cutoffs; Figure 1 ), were 29, 31, 42, and 47 OTUs, respectively.