Bioinformatician, Cellular Biologist, CSO @tychobio_ai @mathurindorel@fediscience.org @mathsrish.bsky.social
Berlin, Germany
Joined September 2016
- Tweets13.5K
- Following1.7K
- Followers424
- Likes82.6K
Pinned Tweet
Midnight reflexion on to improve scientific evaluation beyond h-index: how about a veting mechanism with pagerank based on e.g ORCID profiles.
Each researcher can vouch for any number of other researchers. The importance of a researcher is then simply their pagerank.
I tested waymo for the first time 2 weeks ago and while the gimmick is fun it was supremely annoying that it took longer than public transports because it somehow decided to always take a 10 minutes detour via "waymo highways" with 90%+ waymo cars.
Very herd like.
Results in mice do not translate well to human, especially in ageing!
14% reduction is actually pretty mediocre in mice.
It's a cool scientific result, and definitely worth going through SRSF1 targets to understand the mechanism. Not a lifespan extension cure any time soon.
we might be celebrating birthdays backwards soon
george church's lab just published a single gene that makes old cells act young
- they trained a model to guess a cell's age from its gene expression
- then screened a gene library for anything that made old cells score younger
- the winner was SRSF1, an RNA splicing factor
- gave it to old mice and they lived 14% longer (median), delivered to the liver
- put it in old human skin cells and senescence dropped, DNA repair switched back on
- and the cells stayed skin cells with no identity loss
so aging might be a splicing problem, and a model found the patch
bio/acc
Mathurin Dorel retweeted
Si la DGSI chope les agents du GRU dans les attaques de bâtiments publics , il y faut la jouer médiéval.
Mathurin Dorel retweeted
Ukraine is asking the EU for 70 billion euros to continue fighting in 2027.
That is 0.4% of the EU GDP or 155 euros per European.
Not having a battle-hardened victorious Russian army on Europe's border is worth roughly 2000% of the EU GDP.
Our leaders must find the money.
Mathurin Dorel retweeted
And our big boy (human for size) from @PacBio has now been installed!
Get ready for the highest quality model training data in RNA therapeutics from @tychobio_ai 🧬
Lol opus5.5 for biology. Still not answering questions.
But also: Le Chonk is better that Argon, way to go @MistralAI !
Turns out le Chonk is into rocket science.
Mistral Large 4 scoring well in engineering and astronomy on Ulam’s private benchmark, STEMBench: ulam.ai/stembench
Cool progress.
But the top of the y scale should be 1, because there's still a lot of progress to make.
#SpliceAI2 splice variant effect benchmarks. Considerably improved performance in 9 evaluations spanning cryptic splicing, sQTLs, MPRAs, population genetics, plasma proteins, and rare disease. 🧬✂️🧵
Mathurin Dorel retweeted
For everyone chiming in with "RNA and protein don't correlate"… here's some thoughts I wrote up on this point a while back:
Mathurin Dorel retweeted
Meloni: I do not miss the Russian tanks in Prague and Budapest. I do not miss the Berlin Wall.
If Ukraine stopped defending itself tomorrow, Russia would occupy a European nation and Soviet-style imperialism would return.
Surrender is not peace. It is conquest. 1/
Mathurin Dorel retweeted
Le comportement des États Unis à l’égard de l’Ukraine devrait convaincre les Européens de tout faire pour acquérir leur autonomie stratégique. L’Allemagne peut -elle comprendre que parfois l’indépendance nationale l’emporte sur l’intérêt industriel de court terme?
Mathurin Dorel retweeted
ALL MODELS ARE USEFUL IF YOU KNOW WHERE THEY WORK AND WHERE THEY FAIL. If your hype laden papers & marketing don't have a massive caveats and limitations section listing all the effort u put into finding failure modes, u have not done the requisite work. 16/
Mathurin Dorel retweeted
Nobody in real life cares about ur cherry picked examples of generalization. Show us contexts where it fails. This not only enables users to use ur models effectively but it also helps build credibility in what u are doing & helps u & others push the horizons further faster. 15/
Mathurin Dorel retweeted
1. COMBINATORIAL STOICHIOMETRY: Regulators do not act alone, there is redundancy, competition, synergy & feedback. Knocking them down one by one is insufficient to learn the spectrum of regulatory effects (which is essential to generalize). 2/
Finally, the offer all academics should take!
From the one frontier lab who won't use your data to justify their IPO valuation.
They *will* use your data to train their models, but since they all do it light as well improve the models for science.
We are offering 10,000 scientists the highest usage tier in Google Antigravity free of charge for 6 months. Please apply! forms.gle/WvmBbPvRbpKG1BPn9
Mathurin Dorel retweeted
When we started responding to Russia’s operation against our data centers, and when our strikes reached Yandex’s data centers, it suddenly became a little easier to breathe on many social media platforms. A lot of people noticed it. There was less hate, less division, and fewer provocations.
Because there were fewer bots. And it’s pretty clear where those bots were coming from – if they disappeared when Yandex went offline. This proves, again and again, that Europe is not the source of the problems. Europe is not the one making everything more difficult.
Most importantly, we need people in America, every morning, when the day in Europe is already moving toward evening, to really see what is happening. To see the truth.
From a speech at the Warsaw Security Forum. (3/3)
Mathurin Dorel retweeted
The idea of kicking humans out of scientific research because machines may eventually move faster than humans, and should not be slowed down while human understanding catches up with their progress, is one of the most barbaric, ignorant, dull, and sick ideas I have ever encountered.
It assumes that scientific ideas advance along a linear path toward some absolute, final truth. A conception of science worthy of a bad science-fiction novel.
Science is part of human culture. Kicking humans out of it would not simply be a matter of accelerating scientific progress. It would be a massive transfer of power.
Because if humans cease to participate in and understand the production of scientific knowledge, then whoever controls the machines controls scientific and technological progress, and, ultimately, society.
It is, essentially, an idea subversive of the democratic order.
@TaoistTerence
I fully believe math proofs should be attributed to humans who can explain them.
But this is a very bad take that ignores the transition to Evidence Based Medicine. We do not know why paracetamol, lithium or metformin work, they just do.
We didn't understand how wing worked when planned where invented, humanity didn't understand combustion when fire was discovered.
You can use something you don't fully understand. You'll simply use it better when you do.