Spatiotemporal network structure among "friends of friends" reveals contagious disease process

Geodemographic insights on the COVID-19 pandemic in the State of Wisconsin and the role of risky facilities

The COVID-19 pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) continues to impact the United States. While age and comorbid health conditions remain primary concerns in the community-based transmission of the virus, empirical evidence continues to suggest that substantial variability exists in the geographic and geodemographic distribution of COVID-19 infection rates. The purpose of this paper is to provide an alternative, spatiotemporal perspective on the pandemic using the state of Wisconsin as a case study. Specifically, in this paper, we explore the geographic nuances of COVID-19 and its spread in Wisconsin using a suite of spatial statistical approaches. We link detected hot spots of COVID-19 to local geodemographic profiles and the presence of high-risk facilities, including federal and state correctional facilities. The results suggest that the virus disproportionately impacts several communities and geodemographic groups and that proximity to risky facilities correlates to increased community infection rates.

Social network analysis and whole-genome sequencing to evaluate disease transmission in a large, dynamic population: A study of avian mycobacteriosis in zoo birds

This study combined a social network analysis and whole-genome sequencing (WGS) to test for general patterns of contagious spread of a mycobacterial infection for which pathways of disease acquisition are not well understood. Our population included 275 cases diagnosed with avian mycobacteriosis that were nested in a source population of 16,430 birds at San Diego Zoo Wildlife Alliance facilities from 1992 through mid-2014. Mycobacteria species were determined using conventional methods and whole genome sequencing (WGS). Mycobacterium avium avium (MAA) and Mycobacterium genavense were the most common species of mycobacteria identified and were present in different proportions across bird taxa. A social network for the birds was constructed from the source population to identify directly and indirectly connected cases during time periods relevant to disease transmission. Associations between network connectivity and genetic similarity of mycobacteria (as determined by clusters of genotypes separated by few single nucleotide polymorphisms, or SNPs) were then evaluated in observed and randomly generated network permutations. Findings showed that some genotypes clustered along pathways of bird connectivity, while others were dispersed throughout the network. The proportion of directly connected birds having a similar mycobacterial genotype was 0.36 and significant (p<0.05). This proportion was higher (0.58) and significant for MAA but not for M. genavense. Evaluations of SNP distributions also showed genotypes of MAA were more related in connected birds than expected by chance; however, no significant patterns of genetic relatedness were identified for M. genavense, although data were sparse. Integrating the WGS analysis of mycobacteria with a social network analysis of their host birds revealed significant genetic clustering along pathways of connectivity, namely for MAA. These findings are consistent with a contagious process occurring in some, but not all, case clusters.