It is wild to me that after fifteen years in this field you go to talks and meetings and the standard is still 'let's just find another antibiotic and hope it lasts longer until evolution mysteriously overcomes it' with no sense that it might be an understandable process

I’m in a restaurant listening to a muzak saxophone version of Driver’s License. This is how I know time is accelerating. That sort of metabolism in popular culture used to take a decade

Tying a medical intervention to an ideology (in particular political ideology) has to be one of the most destructive and harmful things you can do

It is the ultimate enemy of evidence based healthcare

Truly excited to see our latest article, in which we were surprised how phages that target conjugative plasmids are both so common and diverse, published!

A real tour de force from Natalia Quinones-Olvera and Siân Owen

nature.com/articles/s4146…

To me, the definitive take on DIY anti-aging was Roy Kishony asking “do you live longer, or does it just feel longer?”

Safe to say this total eclipse beats the last one I was lucky enough to see

Prepubescent ninja turtles?

An incredible piece by Health Nerd. Walking a challenging line that I’ve yet to see someone attempt and, on top of it, do it so well with empathy and science.

“And herein lies the problem with long Covid discussions. There are two separate conversations going on at the same time”

Yes 💯

The fact that others use a term in a different way should not stop you from using it, provided you are clear what you mean when you are using it

Four years on... We should not forget how a pandemic like covid starts, nor that it's a marathon not a sprint

wgbh.org/news/health/20…

Very interesting study. Using a huge cohort and an impressively rigorous design, the authors found that COVID-19 vaccines likely reduced the risk of cardiovascular events by 70-90% for people who caught COVID-19

In short, it depends on who you are...

news.harvard.edu/gazette/story/…)

Yes, distances between points in a UMAP are routinely used to infer biology, even though the distances are meaningless.

Even Leland McInnes, the author of UMAP, doesn't 'put much stock in the exact distances'.

I’d only add the importance of data structure

AI will kill coding in the same way calculators killed math classes. Even if AI allows for plain-language code generation, you still need to know how to structure code to address your particular challenge. The logic, not the syntax, has always been the biggest challenge to learn

beyond eukaryotes have noticed folks using UMAP for things like viruses or bacteria when they could make a more informative and interpretable phylogeny

So agree with 1. And 2. never heard of All of Us before this (and if I were on their leadership team I’d consider rebranding)

Thread on proCov2.

This argument comes up a lot in covid origins discussions -- which covid strain came first?

It's a pity that All of Us used UMAP to visualize ancestry variation in their new marker paper, out today in Nature.
The UMAP algorithm, by design, exaggerates the distinctiveness of the most frequent ancestries, a message that can be misinterpreted by the public.

Let's hope that X can promote this survey to some of the right recipients- sorry to those who were expecting a free house draw or a paywalled newspaper they dont follow...

I ❤️ it when scientists get credit for contributions to the fields they’ve advanced, so delighted to see this celebration of a famed *Structural Biologist* while running

There may of course be other fields in which they have made significant contributions, with less celebration