Open ChristinaB opened 4 years ago
ChristinaB
commented
4 years ago https://twitter.com/UWVirology/status/1241573070269104128
This graphic shows the number of #SARSCoV2 #CoronavirusUSA #HCoV19 cases compared to total tests for each state. Almost 95% of all WA tests have been performed by @UWVirology
@UWMedicine . Read the full report from @politico at https://politico.com/interactives/2020/coronavirus-testing-by-state-chart-of-new-cases/
ChristinaB
commented
4 years ago https://twitter.com/UWVirology/status/1242320571909550081
186 #SARSCoV2 #CoronavirusUSA #HCoV19 genomes have been deposited by our amazing genomics team @UWVirology
@UWMedicine , out of a total of 1111 from the whole world! We agree with the interpretation from our colleagues @nextstrain .
We've just added 100 new Washington genomes, thanks to the incredible work by @UWVirology
@UW and @GISAID . New sequences are in red https://nextstrain.org/ncov?c=recency&f_division=Washington&p=full. The vast majority group within two previously identified Washington clusters, while 8 form separate 1-2 sample clusters.
The most direct connection in the current outbreak, is that the original two strains of COVID-19 have evolved (since early March), and therefore possibly more to be expected (which happened, see Tweets below). We/I don’t know the geography location and history of those testing positive and for which strain. If we could get that data, what else would we need from virology test results (UW Virology or Nextstrain, for example) to understand if the evolution has controlling variables related to geospatial and climate parameters. It is well know that the outbreak is most likely to start where deforestation brings humans into new contact with more risk of viral jumps from animals to humans. (We may need extended land use/land change landscape ontology that doesn't quite exist yet, to make this meaningful at local government and school decision-making scales)
If humans could use Landlab as an experimental platform to study landscape-climate-human relationships to epidemics, that would be amazing. The issue is that the medical records are deidentified such that geohealth models don’t yet connect in a deep or useful way for individual decision making. What I’ve proposed before is to have both educational users modeling with deidentified or synthetic versions of the data, then installing reproducible workflows for privacy protected access.