Thanks to fast sequencing & #opendata sharing from Institute of Viral Disease Control & Prevention, China CDC & @GISAID, 3 sequences from the recent #SARSCoV2#COVID19 Beijing outbreak are now on nextstrain.org/ncov
They fall in clade 20B - frequently observed in Europe
1/6
@GISAID In our global build, the nearest sequences are from the Czech Republic, Taiwan, Greece & Portugal.
However, it's important to remember our builds are downsampled. We can look at the regional builds for different perspectives.
2/6
@GISAID In the Asian build (left) we see same Czech & Taiwan seqs, but also nearby seqs from Denmark, Columbia, Israel, Austria.
In the Europe build (right), we see the Czech & Denmark seqs, but many other sequences from Europe. Some (Portugal, Sweden) are only 1 mutation away
@GISAID The Sri Lanka sequences fall throughout the tree. 3 of them are found in mixed, more European-centered clusters (example in the left image). Only one sequence clusters close to other Asian samples (right image, non-asian samples in gray).
@GISAID@GMalavige@DinukaAri The majority of the new NY sequences falls within a large USA cluster. 4 of the sequences fall separately throughout the tree in more European-centered clusters.
As part of the last push, you might notice things looking a bit different on our website!
As the number of available sequences has grown, we've looked for ways to better display them - so they can be useful within the context that matters, while still giving a global view.
We've moved to a new system where we can do both! We're now running 'Global' builds (nextstrain.org/ncov) which try to show a balanced overview by subsampling down to 120 seqs per division, per month, per year.
2/7
However, to ensure densely-sampled views are still available, we've introduced 'Regional' builds - for Asia, Africa, Oceania, North America, South America, & Europe.
These highlight seqs from that area (color), while putting the global context less prominently (grey).