Assessing the role of live poultry trade in community-structured transmission of avian influenza in China

Yang, Qiqi; Zhao, Xiang; Lemey, Philippe; Suchard, Marc A.; Bi, Yuhai; Shi, Weifeng; Liu, Di; Qi, Wenbao; Zhang, Guogang; Stenseth, Nils Chr.; Pybus, Oliver G.; Tian, Huaiyu

The subnational community structure identified above, which was estimated from the inferred live poultry trade network, provides a framework to evaluate AIV dissemination during outbreaks. However, available AIV genome sequence data from China is not obtained in a structured manner at the national level; hence, few or no genomes are available from some provinces. In order to explore further whether a reliance on convenience sampling hinders our understanding of AIV transmission and spread, we utilized a virus gene flow network (GFN) model that is capable of exploring virus transmission paths through both sampled and unsampled locations (SI Appendix, Fig. S5 ). The GFN model generates measures of virus gene flow among locations that can be directly compared with the poultry trade network reported above. We find that the virus gene flow network closely matches the structure of the live poultry trade network (i.e., Fig. 3) . In both the live poultry trade network (SI Appendix, Table S2 ) and the gene flow network (SI Appendix, Table S3 ), we identified provinces that acted as hubs (Anhui, Henan, Hubei, and Hunan) by calculating the in-degree and outdegree of nodes. These hubs, located toward the center of the country and linking adjacent communities, may play a significant role in virus transmission among the five region-level communities. However, the structure of the virus gene flow network (SI Appendix, Fig. S5C ) is notably different from that of the poultry egg trade network (SI Appendix, Fig. S6 ).