High-throughput protein design and biophysics | Asst Prof, Northwestern Pharmacology @NuFeinbergMed @NUSynBio.
Chicago, IL
Joined August 2008
- Tweets1.5K
- Following611
- Followers3.1K
- Likes5.7K
Pinned Tweet
New preprint! We measured temperature- and pH-induced aggregation for over 18,000 natural and de novo designed protein domains!
Gabriel Rocklin retweeted
DO NOT BE FOOLED BY THE "SUPPORT SCIENCE" PROPAGANDA. If you live in California VOTE NO ON PROPOSITION 38 and tell everyone you know to do the same. It will not support science! It's a terrible way to fund research that, among other things, earmarks billions of dollars for one institution.
Gabriel Rocklin retweeted
Avengers assemble: 1 year BakerLab + NVIDIA GenAIR = 🔥 rfd4-proteina.org
Presenting RFD4-Proteina, our latest protein design model, significantly better and faster than everything we compared against, both in silico and in the lab! 1/30
Gabriel Rocklin retweeted
We all want more measurements with faster turnaround, but there is a drawback to automating labs that is not often discussed: it will stymie innovation.
Here’s why: … /1
I know you love your transportation analogies (no dig - genuinely useful), but here Id offer that we have more of an iterative search problem. Science is NOT agile. The latency between experiments (especially if you change a few things or decide to change attack vectors) is huge.
So what can we do to reduce that latency? That’s more than just “run it through faster” (though that’s level 1 for sure), it’s also “what stuff do I need?”, how many steps back do I have to go (can’t prepare for all eventualities - combinatorial explosion).
I need approaches people - where’s the wisdom of the crowd? Can’t be just me and @jrkelly talking to each other (though always nice).
Get after it!
Let me know! Let me know I’m full of it!
Gabriel Rocklin retweeted
I have postdoc positions available to lead two new exciting projects! One is on in situ protein sequencing (collab with Ed Boyden at MIT) and the other is on modeling cryptic pockets for drug discovery (collab with Scripps and MGH). Reach out if interested! Details below
Come join!
The Molecular Biosciences Department at Northwestern is excited to announce a faculty search for a Tenure-Track Assistant Professor position!
Learn more and apply at: molbiosci.northwestern.edu/f…
#BiologyJobs #AcademicJobs #FacultySearch #FacultyJobs
Gabriel Rocklin retweeted
🚀 Interested in joining my group as a PhD or Postdoc?
I'm mentoring for the ETH AI Center's Doctoral and Postdoc Fellowship, working on biomolecular design.
If you're excited about AIxBio, I'd love to hear from you.
Apply now 👉 ai.ethz.ch/apply
Deadline: Oct 27, 2026
Gabriel Rocklin retweeted
Thrilled to announce that the Savinov lab is now live at @UTAustin! Our lab @UT_MBS employs massively parallel experimental and AI methods to discover and design protein fragments as universal regulators of protein interactions in health and disease. savinovlab.com/ (1/n)
Gabriel Rocklin retweeted
We are hiring! 🌟 Dept of Biochemistry and Molecular Biology at the University of Chicago is searching for a tenure-track assistant professor. It's a broad search: we want great colleagues doing serious biological work at the molecular scale. Please share!
apply.interfolio.com/191981
Gabriel Rocklin retweeted
Rents are declining in 2026 because of boom in apartment construction in 2024. 600,000 rental units were built, the most since the 1980s. Skill-testing question: who was President in 2024? (Need a hint? OK - not the same guy who was president during COVID lockdowns of 2020.)
JD Vance: "Rents are falling for the first time in a decade under the leadership of Donald Trump because we got people who shouldn't be here the hell out of our country"
Readers added context they thought people might want to know
Rents in the US are rising, not falling.
ipropertymanagement.com/research/avera…
fred.stlouisfed.org/series/CUUR000…
Gabriel Rocklin retweeted
“Maximum transparency in all that we do”.
One of the biggest things that they “do” is to select some grant proposals for funding and to reject the rest.
Will he be transparent about this process?
Strict payline ICs are in a sense the most transparent. He’s halting that.
NIH Director Dr. Jay Bhattacharya has announced that the NIH Public Access Policy, originally slated to go into effect on 12/31/25 will now be effective 7/1/25. Read more in his statement: bit.ly/4jTQI4O
ALT Graphic with blue background, HHS and NIH logo, and a photo of NIH Director Dr. Jay Bhattacharya. Quote included: "Providing speedy public access to NIH-funded results is just one of the ways we are working to earn back the trust of the American people. Trust in science is an essential element in Making America Healthy Again. As such, NIH and its research partners will continue to promote maximum transparency in all that we do."
Gabriel Rocklin retweeted
CSB's DREAM is hiring. Interested in joining the team? Explore open positions here: syntheticbiology.northwester…
Gabriel Rocklin retweeted
We have a new paper out on bioRxiv describing how to turn "static" small-molecule binders into dynamic shape changers! Short 🧵
biorxiv.org/content/10.64898…
Gabriel Rocklin retweeted
ProteinGuide is finally officially out. Congrats to the amazing team of @junhaobearxiong, @Ishan__Gaur, @MariaLukarska, and Hunter Nisonoff, in our continued great collaboration with @SavageCatsOnly.
Property guidance for protein sequence generative models with ProteinGuide nature.com/articles/s41587-0…
Gabriel Rocklin retweeted
We show which mutations are rescued using the published tool compound YC-001. Internally at Octant, we use the same methodology to run deeper dose-response curves across our clinical candidate OCT-980 to estimate EC50's of every possible mutation, helping us design precision trials. 5/
Gabriel Rocklin retweeted
Please RT: We're hiring! Computational postdoc for foundation model and sequence-structure-function studies of intrinsic disordered proteins through a consortium and unique large datasets. Ideal candidates have led ML/AI projects in protein design/analysis.
Cool! And pretty sure the experiments here come from @KotaroTsuboyama !
Gabriel Rocklin retweeted
LLMs show strong biomolecular intuition without seeing alignments or structure:
Here, Claude Opus 5 inferred a correct α–β–β–α topology from only the sequence and mutant variants, placing all four elements within 1–3 residues of the withheld experimental structure (ρ = 0.68).
Gabriel Rocklin retweeted
Excited to share my first co-corresponding preprint in collaboration with @realJimWells and Bill DeGrado, where we used de novo protein design to develop a photocatalytic proximity labeling enzyme EYClamp. biorxiv.org/content/10.64898…
Gabriel Rocklin retweeted
on the one hand: the actually existing trump administration, which gave RFK jr a cabinet level post that he used to do things like cancel mrna research in favor of instagram essential oils
on the other hand: these keys! so shiny! they go jingle-jangle! haha! I love them!
