Global patterns in coronavirus diversity

Anthony, Simon J.; Johnson, Christine K.; Greig, Denise J.; Kramer, Sarah; Che, Xiaoyu; Wells, Heather; Hicks, Allison L.; Joly, Damien O.; Wolfe, Nathan D.; Daszak, Peter; Karesh, William; Lipkin, W. I.; Morse, Stephen S.; Mazet, Jonna A. K.; Goldstein, Tracey

Certain limitations should be considered in the interpretation of our data. First, the diversity of CoV sequences detected is almost certainly biased by our sampling effort. A small number of species were sampled abundantly, thus maximizing our chances of finding a CoV (27/282 species had >110 individuals); however, most were under-sampled (232/282 with 50 individuals). While this biases the probability of finding a CoV towards the more abundantly sampled species, we stress that it also reflects the uneven distribution of naturally occurring communities (Magurran 2011) , and that targeting the very rare species would be prohibitively expensive. It is also unclear if population size has an effect on the number of viruses present in a species-e.g. do larger populations accommodate more viral diversity? Second our analysis only represents the 'last' evolutionary event identified. It does not represent, or account for, the complete evolutionary history of these lineages. Therefore, while a single event has been ascribed to each association (see 'Methods'), it does not preclude other events having an equal or greater impact in the past. We are limited to using this most recent event type in order to assign a single evolutionary event to each unique association. It should also be noted that events are dependent on the relationships observed in the reconstructed trees and could change as more viruses are added or longer sequences are used. Third, we have only generated partial sequences for analysis, which limits our ability to comment on the specific zoonotic potential of each virus or provide a comprehensive analysis of their genetic histories and taxonomy, for example, estimating the frequency or impact of recombination, which can be an important driver of host switching and disease emergence (Anthony et al. 2017) . We certainly support fullgenome sequencing wherever possible [and have such efforts underway (Anthony et al. 2017) ], however difficulties in virus isolation or sequencing directly from swabs (where material and viral load can be low, yielding variable results) can often preclude extending sequences much beyond the short fragments amplified by consensus PCR (Drexler et al. 2014 ). Equally, logistic and permitting issues in moving biological samples across borders, and the need to first develop in-country capacity to fully sequence these viruses, have all limited our ability to characterize (most of) these viruses further.