Tim Stearns (@StearnsLab) header
Tim Stearns (@StearnsLab) profile picture

Professor and Dean at The Rockefeller University. Cell biologist. Believer in the power of science education.

New York, NY
Joined November 2010
Photo by Tim Stearns (@StearnsLab)Photo by Tim Stearns (@StearnsLab)Photo by Tim Stearns (@StearnsLab)Photo by Tim Stearns (@StearnsLab)Photo by Tim Stearns (@StearnsLab)Photo by Tim Stearns (@StearnsLab)Photo by Tim Stearns (@StearnsLab)Photo by Tim Stearns (@StearnsLab)Photo by Tim Stearns (@StearnsLab)Photo by Tim Stearns (@StearnsLab)
Adding US federal budget, with the two curves indexed at 2003. The NIH budget has been flat for 20 years, and is a decreasing fraction of federal spending (the big bump in 2020 was pandemic spending).
The US NIH budget, our crown jewel for advancing biomedicine, is the same as it was over 20 years ago, inflation adjusted
2
7
21
2,718
Tim Stearns retweeted
NYC Postdocs: Join us next week at NYC Postdoc Night Science at CUNY. I'll be there as well as well as @StearnsLab to exercise tools of the creative process. Come and meet other amazing NYC postdocs like @KelseyRMonson and @rmassonix . Thank you Jaeseung Hahn for organizing this! Free registration: lnkd.in/eAJzbR9t. For more on Night Science: night-science.org
2
2
4
1,691
Tim Stearns retweeted
Please share: The Center for Studies in Physics and Biology at The Rockefeller University is hiring postdoctoral fellows. Apply to join our community. Application here: cspbfellow.rockefeller.edu/
10
1
9
799
NYC Marathon today! Men's leaders, just after crossing the Queensboro bridge, with 10 miles to go. Almgren went on to win.
1
10
1,939
The Guggenheim in the infrared
2
11
612
Lots of witches and goblins out in NYC today, but you really can’t beat the 1500’s for the macabre. “Salome with the Head of St. John the Baptist” 1507 by Andrea Solario, at @metmuseum
1
5
787
The field of cell biology was born at Rockefeller thanks to Albert Claude, George Palade, and Christian de Duve, who shared the 1974 @NobelPrize for their discoveries on the structural and functional orga­nization of the cell. #125YearsofRockefeller
2
7
29
3,039
solid foundation in statistics is even more important in the age of AI. This interactive app explaining statistic concepts is great seeing-theory.brown.edu/inde…
2
67
442
26,179
Great blog post from Matthew Schwartz about the power of working with LLMs to solve “Claude-shaped” probldems. Figure out what the tool is good at (and they are *really* good at many things science) and exploit those, with a large dose of human expertise. anthropic.com/research/claud…
6
916
Tim Stearns retweeted
🧵 Biohub, @ENERGY, @NIH, and new funding partners today announced a major expansion of an international effort to generate and make accessible the data enabling predictive AI models of biology. Together, the organizations are investing $1.8 billion in funding, data, computation and new measurement technology - largest coordinated commitment to map the biology of the cell to power development of digital models of biology. bit.ly/3UhjoxD
24
102
46
422
82,656
Congratulations to Karl, Peter, and Georg! Another great example of how basic research - in this case, understanding how an algal cell responds to light - can have an enormous impact on our understanding of something as complex as the brain.
BREAKING NEWS The 2026 #NobelPrize in Physiology or Medicine has been awarded to Karl Deisseroth, Peter Hegemann and Georg Nagel “for their discoveries concerning light-gated ion channels and optogenetics.”
4
9
62
3,514
Rockefeller grad student @NatJAlexander of @ElaineFuchsLab investigates the connection between #obesity and #inflammation in skin—a focus she crafted by following her curiosity through different labs and scientific disciplines. 🔗: bit.ly/3VxJSuW
1
6
17
1,651
Great job pulling back the curtain on the polyclonal antibody industry. Fake validation, poor quality antibodies, absurdly high prices, and it has been like this for decades. See paper by Kahn, et al. for a way forward: elifesciences.org/articles/1…
A few months ago I was browsing google images looking at p53 western blots in the SK-OV-3 cell line for an unrelated project. I noticed an unusual western blot hosted on Thermo Fisher's website which I posted on social media. Usually I just post weird blots to PubPeer. 👇
1
20
133
15,492
In my experience most people with "magic hands" at the bench actually just had a better understanding of their experimental system, and how experiments work. Harder to reproduce that in the roboticization of lab science. nytimes.com/2026/08/27/scien…
13
15
2
171
20,866
I’m not saying that what’s described in the article isn’t important - it is - but rather that there are other variables with larger effect sizes.
1
8
807
At the wonderful Zoological Museum in Palermo, Sicily, contemplating the extraordinary developmental plasticity that allows one eye of a flatfish to migrate from one side of its skull to the other during metamorphosis.
5
658
The perfect species name doesn't exi...
2
633
96 years old and still the best spire in NYC.
2
20
954
Tim Stearns retweeted
🔥The NIH has funded our R25 grant to bring Night Science training to all 40,000 US postdoctoral fellows! It all started with discussions at the Radcliffe Summit we organized 2 years ago at Harvard where the question was how bring creativity training into the sciences. Our R25 team includes the amazing @StearnsLab, @brangwynnelab, @OylerYanivLab, @JamesKaufman, Maura N. Polansky, Oliver Bogler and @MartinJLercher.
17
25
303
20,225
AI has helped resolve an important question in statistics. In the area of multiple hypothesis testing, the goal of controlling the false discovery rate (FDR) has been introduced in a seminal paper by Benjamini and Hochberg (1995). They also introduced a method (the Benjamini-Hochberg or BH method) and proved it controls the FDR. This method has been widely adopted in modern high-throughput science, including in genomics, astronomy, economics, etc. The paper has has garnered more than 130,000 citations to date. However Benjamini and Hochberg showed FDR control only when the data for the individual tests are *independent*. In practice, these data are often dependent; a good example is data on genetic variants due to linkage disequilibrium. Later work has focused on extending the validity of the BH procedure, e.g., to a form of positive dependence by Benjamini and Yekutieli (2001). The question of when the BH procedure controls the FDR has remained open. Over the last twenty years, many authors, including Reiner-Benaim (2007), Kim and van de Wiel (2008), Benjamini (2010), Sarkar (2023), Sarkar and Zhang (2025), have conjectured that the BH procedure controls the FDR for two-sided tests using any correlated Gaussian data. These authors have presented both theoretical and empirical evidence supporting, but not directly showing, the conjecture. With the help of AI (specifically GPT-5.6 Sol Pro), I have settled the question in the negative: The Benjamini-Hochberg procedure does *not* generally control the false discovery rate at the desired level for correlated two-sided Gaussian tests. This was done by exhibiting a Gaussian factor model for which, at a nominal level alpha=0.01, the false discovery rate is proved to be FDR>0.0104. There is a lot of interesting commentary to be made: 1. This result should be of interest to everybody in the field of statistics. Emmanuel Candes of Stanford University once called the false discovery rate and the Benjamini-Hochberg procedure "one of the two most important developments in statistics after 1950" (the other being James-Stein shrinkage). The present conjecture is probably the most central question about FDR/BH that was unresolved to date. 2. GPT-5.6 one-shot the problem after 90 minutes of reasoning, whereas with 5.5 I was not able to solve it even after iterating with multiple parallel agents for perhaps 20 hours. So the capability improvement is quite real. Exciting times to live in! 3. The argument is not especially surprising, but it does combine an asymptotic approach (standard for FDR analysis, see e.g., Genovese and Wasserman, Efron, etc) with a numerical certificate in a way that would be pretty non-standard in the field. Once we have the specific example, then straightforward simulations also support that the false discovery rate is indeed higher than the nominal value (see attached fig). 4. The current degree of violation over the nominal level is relatively small (0.104 vs 0.1). So the importance of this result is mainly conceptual. The practical implications remain to be determined. Overall, an exciting development! Preprint is available here (faculty.wharton.upenn.edu/wp…) and will be on arxiv tonight; supporting code is here (github.com/dobriban/BH).
77
341
69
2,000
1,153,922