🎉VERY excited🎊 to share news of the shiny new #monkeypox@GenSpectrum (previously CoV-Spectrum) website & API - allowing lightning-fast ⚡️ queries to get info on metadata & sequence information.
As more #opendata sequences come in, this will be incredibly powerful!
I hinted at this coming a little earlier today 😉 - but want to really emphasize: this is what truly #opendata looks like.
Enabling development of tools that are able to benefit everyone to do fast & effective research to underpin response & understanding
On the other hand, it's critical to recognise that #opendat is only possible if generators are able to be appropriately credited for their work, & thus feel safe sharing it. We need more effort, particularly from journals, to make sequence scooping for publications history.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
This morning's session at the #Nobel Symposium starts with Michael Farzan explaining ACE2 binding in #SARSCoV2 & #SARSCoV1, including a great timeline comparison between how long steps took between the two (much faster in SC2) & asking if we can improve vaccines.
@MichaelWorobey follows, speaking to the importance of digging into the evidence of #SARSCoV2 outbreak sources & Hunan market.
He opens with a remark about how this beautiful location with so many great people to talk to is pretty close to heaven - he's not wrong!
Hyeryun Choe closes the first session with a great talk on how #SARSCoV2 adapted to humans, and the importance of the D615G mutation, to stabilizing the protein. With a great graphic reminder of how quickly G614 dominated.
Feels very timely today to be on the way to the #Nobel Symposium in Stockholm: The First Great Pandemic of the 21st Century.
Though I do not believe the #monkeypox outbreak will be a pandemic, it's a reminder there are many pathogens that can rear their heads in different ways.
Tomorrow, I'll talk about how phylogenetics played a critical role in #COVID19#SARSCoV2 - but also outline what challenges we face when we look to tomorrow's pandemics. In particular we have to ensure: sequencing, sharing, software & sufficient background.
A lot of pandemic preparedness focuses on detecting a new pathogen, or what to do once we know it's bad news. But a lot of the critical work is a little more boring: encouraging #openscience, ensuring credit, maintaining & advancing software & building up our virus knowledge.
You can find these links in a dedicated box, just above the gif protein structures - just click on any of the available proteins (any with mutations) to go straight through to Aquaria.app!
2/4
For details on how this works, check out @aSchaFu's fantastic thread, including some of the amazing things you can investigate using Aquaria.app:
In line with the recent addition of 22A (BA.4), 22B (BA.5), & 22C (BA.2.12.1) as @nextstrain clades, you can now find these three variants on CoVariants.org!
As usual, each new variant has their own page, where you can see defining mutations (left), a per-variant plot (bottom) and read about the variant, its mutations, & studies done so far!
Note that these are *estimates* of the proportion of cases by variant. Only a small % of cases are sequenced - we're extrapolating, so interpret with care!!
2/7
Currently we only show countries that have sequenced at least 3% cases for >=50% of the 2-week period CoVariants tracks (since May 2020). This tries to avoid plotting data that may not represent cases well.
If you think a country has good data & is missing - let me know!
3/7
CoVariants now filters out more data that can lead to misleading or confusing variant trends! Thanks to eagle-eyed users 🔎 & @MoiraZuber!
1/5
1 - CoVariants now only displays variant data from sequences that have come from humans.
A variant may circulate longer in animals than in people, & focused sampling of those animals can thus give a misleading picture. See mink EU1 samples in Latvia, before & after!
2/5
2 - Many people had flagged strange pockets of recent 'other' sequences in countries like Italy, India, Chile & others.
Was this a new variant?
No: most of these are just bad sequences where a clade can't be assigned! CoVariants now filters these out entirely!