Профиль

So the 2026-05-25 Ontario sequence is much more closely related to the other five than it appears. Two identical branches (both with G22882T and G29779T) get artificially separated here. @angieshinrichs.bsky.social, is there a way to get G22882T and G29779T included in the PJ.2.1 Usher branch?

Yes, and both very closely related to recent Ontario and New York sequences. Usher masks S:K440N, which actually occurred together with S:N439K, H445P, and W452K on the far left of this tree. It also masks G29779T, which is in all PJ.2.1 but gets tacked on at the end.

You'd probably find this interesting, @solidevidence.bsky.social.

She got into a very bitter, ugly, and public Twitter battle with AJ Leonardi and seemed a different person after it. That whole affair is a cautionary tale about engaging in social media feuds.

I remember playing a lot of Ocarina of Time that winter.

You could walk on top of the snow because the ice layer on top of it was so thick. And it was so cold that salt couldn't melt anything, so the roads were covered with an inch of solid ice for almost 2 weeks. We were out of school for a month: 2 weeks Christmas Break followed by 2 weeks canceled.

That year in NE Indiana, we never had winter. I was in 8th grade. Every year, the 8th grade at our school went on a skiing trip. Postponed our trip like 4 times before finally canceling. Too damn warm. The next year we got 3/4" of freezing rain on top of tons of snow along with freezing temps.

Is this not an indication of lower ascertainment? Unless the virus is getting inherently more virulent as it transmits, this seems likely to mean that as time goes on, only the most severe cases are getting tested and detected, with less severe cases going unconfirmed. Seems like a terrible sign.

Another finding: The number of S1 mutations in lineages predicts fitness at least as accurately as machine learning and fancy protein-language models. Here's a link to the interactive version "the intended reading experience of this paper." blab.github.io/fitness-flux/ 2/2

Haven't seen this shared, but it's a great one by a giant in the field, Trevor Bedford. Finds that fastest SARS-2 fitness growth was 2021 to mid-2022. Since then there's been a steady rate of evolution ("fitness flux") >4x faster than H3N2 flu A, a benchmark for rapidly evolving RNA viruses. 1/2

bsky.app/profile/ryan... Phylogenetic tree view below. Still too soon to make out what it's overall growth rate is like. 2/2

Visual summary of PJ.2.1'sAA muts on top of BA.2.86. As @solidevidence.bsky.social shows, it's spread quite far geographically, from its origin in Ontario, Canada, to (at least) Alabama, Connecticut, Florida, Hawaii, Maine, New York, Utah, Vermont, Virginia, & DC. I may be missing some. 1/2

"All of these [AI executives] know there's an economic reckoning looming and can feel the heat of the public backlash on the back of their necks. As they get increasingly desperate to keep the public distracted, expect the misrepresentation to only get dumber and more dramatic." 2/2

This is excellent. "The 'rogue AI' narrative helps obscure some pretty ugly truths.... Like the fact that OpenAI lost $38.5 billion last year, will likely never make a profit, & is facing total collapse or acquisition as an AI bubble looms." Also, a vision of an even dumber near future: 1/2

Not sure if it made it's way here, but yesterday on Twitter there was this farcical attack on Fauci because in March 2020 he said HCQ might be tried & that we should be cautious given the unknowns about remdesivir—boring statements any reasonable person could've made at the time. >1.6K retweets

Loved that guy. He did a split single with Jason Molina back in the day, and after Molina died, he did a 5-song EP of Molina songs as a tribute, including two songs from Molina's self-titled debut Songs:Ohia album. To me, this is the greatest thing he ever did though. bsky.app/profile/ryan...

RIP Glen Hansard. Such a great singer & artist. Generally, I think it's impossible to cover Neutral Milk Hotel and do the songs any justice, but Glen, with the Swell Season, did this unreal cover of "Two-Headed Boy" 15 years ago. It still gives me chills. www.youtube.com/watch?v=-EPZ...

There are different types of CADR though. Some companies only report one rate without saying whether it's the smoke, dust, or pollen CADR. My assumption is that with a non-HEPA filter the smoke CADR is lower than with a HEPA filter. Presumably companies reporting just one CADR are reporting pollen.

There is a good system for naming variants, but these are names for individual sequences. Part of the problem is that GISAID, the largest database, is run dictatorially by a con man with no background in science or public health and who does everything he can to undermine open data sharing.

Similarly, with Michelle Obama's healthy eating initiatives: It was clear there would be no substantive action, like taking on food corporations, changing farm subsidies, or restricting junk-food advertising. Instead, we got empty moralizing & counterproductive restrictions on school lunches.