Joined July 2022
- Tweets 1,572
- Following 211
- Followers 2,528
- Likes 845
302 Photos and videos
Prof. Michael Lin retweeted
🍹Weekend project: I (with help from claude - budget $40), was able to get OpenFold3 weights running inside AlphaFold3 codebase (jax). Works for proteins, ligands, rna/dna (1/3)!
AF3 source code is now open source (aka Apache 2.0)! 😎
Though not the weights... 🙃
github.com/google-deepmind/a…
3
39
241
12,047
Prof. Michael Lin retweeted
Jun 13
A terrific perspective from Ivan Dikic
linkedin.com/feed/update/urn…)
2
16
65
22,370
Prof. Michael Lin retweeted
Jun 11
Over the past two years, Genentech has laid off or lost Ira Mellman, Shannon Turley, and now Vishva Dixit!
What is happening at Genentech?
Jun 11
SCOOP: Roche’s Genentech underwent another round of layoffs this week in its early research and development group, Endpoints News has learned, which included shuttering two research units and the departure of several senior researchers.
My latest here: endpoints.news/exclusive-gen…
11
13
105
51,940
Prof. Michael Lin retweeted
🧵 Thrilled to share our new paper, just out in @NatureComms! 🎉 BTSP is known to rapidly form place cellsin hippocampal CA1 via plateau potentials. Here, we found that BTSP isn't just for spatial maps. It works for non-spatial information too! 🧠⬇️
📄doi.org/10.1038/s41467-026-7…
4
16
58
4,414
Prof. Michael Lin retweeted
✨Join us for GCNI seminars:
🗓 Date: June 25, 2026
🎤 Speaker: Prof. Andreas Schaefer - The Francis Crick Institute
🗓 Date: June 26, 2026
🎤 Speaker: Prof. Michael Lin - Stanford University
Can’t wait to see you there! 🎉
lnkd.in/ep5YMBur
1
2
291
Prof. Michael Lin retweeted
Jun 9
At some point we have to ask ourselves: what are we doing here? I am old enough to remember when an R01 was sufficient to run a mid-size lab. Now it's effectively a postdoc fellowship. I don't think that is healthy for science. At some point, we have to make some hard decisions.
Jun 9
An R01 is around $250k (modular) to ~$500k. I just wrote an R01 that included NO FACULTY SALARIES (or we only asked for about ~1 month effort each) with 2 other PIs and we each had about 1.5 students on it.
11
11
184
37,906
Prof. Michael Lin retweeted
Jun 8
Antibody LMs learn what looks antibody-like, but not how selection turns naive germline antibodies into strong binders.
@aakarshv1 and I are excited to share CoSiNE, a model that learns this germline-to-mature process for variant effect prediction and antibody design. (1/8)
8
50
221
41,577
Prof. Michael Lin retweeted
So proud of this paper led by Conor Dorian, in collaboration with Jiannis Taxidis and Ahmet Arac where we used 2P opto and calcium imaging to show BTSP creating non-spatial representations in hippocampus during working memory task performance. nature.com/articles/s41467-0…
1
9
45
4,286
Prof. Michael Lin retweeted
What nonsense. Successful deployment of science in private industry has origins in academia. Continues to be the case today. There r big problems with academic incentives & there are ways to reform this constructively. I am actively advocating for positive change. Not this BS.
3
45
3,953
Prof. Michael Lin retweeted
Jun 9
In other news, today the NIH proposed caping the maximum of grants at 2 per research lab, including collaborations.
grants.nih.gov/grants/guide/…
We're living in the upside down.
Jun 8
Lila Sciences may raise about $2 billion at an $8.5 billion valuation.
That number looks incredible if you treat Lila as a normal AI biotech. It makes a little more sense (maybe!) if you treat it as a bet on a new R&D operating system.
Lila’s pitch is AI Science Factories: AI systems connected to automated labs that generate hypotheses, design experiments, run them, learn from the results, and repeat. In other words: model, robot lab, experiment, data, repeat. The whole shebang.
Or “scientific superintelligence” and “operating system for science," as Lila calls it. (3/7)
14
38
195
73,281
Prof. Michael Lin retweeted
SCI member Michael Lin & others developed MIDAS, a rapid protein-engineering platform that could accelerate cancer drug & diagnostic development by helping researchers quickly build & test thousands of protein variants in mammalian cells. #Bioengineering brnw.ch/21x3blA
1
3
468
Prof. Michael Lin retweeted
I just completed my service in my NIH study section. Some thoughts - with the disclaimer that these are personal, subjective impressions and do not reflect the opinion of any official body. Other than the one attached to my head. (1/10)
7
26
150
45,614
Prof. Michael Lin retweeted
I am sending an open letter to Thermo Fisher. Their response to my response to their manipulated western blot is bullying and petty. Yes, this western blot really is manipulated, it is unfair on me to say otherwise. I don't make those accusations lightly. #ThermoFishy
22
73
514
51,484
Prof. Michael Lin retweeted
Jun 5
What if a binder could not only bind a target, but selectively recognize one conformational state over another? Today, we’re excited to share AlloGen, our experimentally-validated framework for conformation-selective binder design! 🐙
📜: arxiv.org/abs/2606.05474
🤗: huggingface.co/ChatterjeeLab…
🧵👇🏾
6
55
294
34,528
Prof. Michael Lin retweeted
Thrilled to share our latest publication in @PNASNews —and the first publication from my lab!
pnas.org/doi/10.1073/pnas.26…
2
17
76
8,690
On my other handle, I read a lot of people denigrating those with healthcare experience as dinosaurs whose expertise is not needed
On this handle, I read a lot of ML people denigrating those with experience in scientific experimentation as dinosaurs whose expertise is not needed
1
10
1,313
It's ironic that ML is being used as a way of devaluing experts, the same ones that actually created new insights and will continue to do so
5
424
Finance guys will back whatever is hot right now
I've been in this field 30 years now and have only seen a steady erosion in what was considered long-term investing: from 10 years to essentially 1 year
Jun 3
I need to start my own VC fund of scientist backing scientists instead of finance guys backing whatever’s hot right now.
7
1,609
Also why tech bros think LLMs can solve scientific questions
Jun 3
Most experiments fail, and negative results rarely get published. This means LLMs are unaware of the outcomes of most experiments.
4
3
41
6,199