iiinova — where your digital self keeps evolving

It's horrifying to see how the admin and DOGE have intentionally decimated the NIH through layoffs, freezes, and multi-year funding (a stealth budget cut)- so that NIH grad students have to moonlight as admins to get grants out bsky.app/profile/maxk...

This is just a short excerpt for the contact page.

This early NIH grant was a huge boost to my new (at the time..) lab- it's how we recruited @ajrferrari and many others. I shared the grant in the docs repo of New PI Slack - happy to share if you'd like newpislack.wordpress.com/contact/

newpislack.wordpress.com

Join

NIH Director's New Innovator Award - Funded Research

Our project was funded by the NIH high-risk high-reward program, and it was exciting to pitch it to a broader audience and highlight the long-term potential for ML and protein design- even before there was AlphaFold 2! commonfund.nih.gov/newinnovator...

commonfund.nih.gov

Funded Research

121

Simulations can be valuable but there is a huge need for scalable experiments that probe protein dynamics- of course we're expanding our work here and there's also great work from others @hkws.bsky.social @ginaelnesr.bsky.social Basile Wicky (not on bsky) @diffuseproject.bsky.social

151

Relying exclusively on MD simulations to build dynamics into ML models runs some risk of garbage in/garbage out- simulations are not "gold standards" (our experiments in D2O aren't perfect gold standards either for that matter!). Excited to benchmark/improve sims with exp data!!

It's amazing to see explosive growth in machine learning/modeling of protein structures, but we have so far to go to accurately model/incorporate dynamics and fluctuations. Especially because these happen at many scales- from rotamer shifts to global unfolding and all in between

121

I also want to especially thank Mike Guttman at UW for sitting with me for hours/days to train me in HDX-MS even though I was a postdoc in a different lab with a weird idea and he was just starting a faculty position. It was super generous

Gabriel Rocklin @grocklin.bsky.social · 13/05/2026

thank you! and thanks for all the inspiration on this project!

Kresten Lindorff-Larsen @lindorfflarsen.bsky.social · 13/05/2026

Reposted by Gabriel Rocklin

If you missed the preprint, now is a good time to read the journal version of @grocklin.bsky.social et al’s fantastic paper on multiplexed HDX measurements Large-scale discovery, analysis and design of protein energy landscapes doi.org/10.1038/s415...

13911

Reposted by Gabriel Rocklin

Alondra Nelson @alondra.bsky.social · 18/04/2026

Another example demonstrating that the federal science and research ecosystem is being seriously damaged regardless of what’s ultimately appropriated

5463179

Reposted by Gabriel Rocklin

Calin Plesa @calin.bsky.social · 03/03/2026

New lab preprint: if ML struggles with extrapolation, let's expand the diversity of training data with gene synth, DNA shuffling, and ML gen.... also lots of #FluorescentProteins #ProteinEngineering #MachineLearning #SyntheticBiology 1/n

2135

Reposted by Gabriel Rocklin

bioRxiv Synthetic Biology @biorxiv-synthbio.bsky.social · 03/03/2026

High Diversity Gene Libraries Facilitate Machine Learning Guided Exploration of Fluorescent Protein Sequence Space www.biorxiv.org/content/10.64898/20…

032

Reposted by Gabriel Rocklin

Nick Polizzi @nickpolizzi.bsky.social · 20/02/2026

I’m looking to hire a research technician for my lab at Harvard & DFCI, who would primarily work in the wet lab expressing and characterizing designed proteins, starting this summer. A great role for a recent college grad looking for an immersive research experience before grad school.

11113

Reposted by Gabriel Rocklin

Keating Lab @keatinglab.bsky.social · 18/01/2026

Here’s a preprint from the Keating Lab: www.biorxiv.org/content/10.6.... Foster Birnbaum and Amy E. Keating demonstrate that sequence design models, such as ProteinMPNN, are limited because they were trained only on native sequences.

biorxiv.org

194

Gabriel Rocklin @grocklin.bsky.social · 20/01/2026

Yeah this seems terrible. On the other hand, if funding levels stay ~constant, I wonder if this will be undone in the future, leading to a kind of random jump up in the number of projects supported for a few years (far above even 2024) in ~2030? Not that that's much to get excited about.

Gabriel Rocklin @grocklin.bsky.social · 16/01/2026

Nope, just trying to force some honesty from the side that introduced a very dishonest mechanism of cutting budgets. Fruitless I know

010

Gabriel Rocklin @grocklin.bsky.social · 16/01/2026

I know negotiating with a bunch of bad-faithed liars can be a waste of time, but has anyone proposed a compromise where we switch to MYF but allocate extra funds so that the transition is net neutral to the # of awards? Seems like that would clarify what we're supposedly debating here.

200

Gabriel Rocklin @grocklin.bsky.social · 07/01/2026

Gonna be funny when people get rejected for being at the wrong career stage (like the K99)- “PI is WAY overqualified for retirement”. “It is not clear PI will benefit from additional research prior to retirement”. “PI not qualified for retirement, hasn’t graduated enough students to be competitive.”

0120

Reposted by Gabriel Rocklin

Michael Li (李之樸) @mcpli.bsky.social · 31/12/2025

The final Calvin and Hobbes, which appeared in papers 30 years ago today.

122130424237

Reposted by Gabriel Rocklin

AI x Bio Discovery @aixbiobot.bsky.social · 29/12/2025

Large scale prospective evaluation of co-folding across 557 Mac1-ligand complexes and three virtual screens [new] Mac1 co-folding eval: Pose prediction, conform. change, & hit ID post-training.

053

Reposted by Gabriel Rocklin

Katie Mack @astrokatie.com · 30/11/2025

The most precious commodity you have is your attention. You don’t have to waste it on poor-faith debates or arguments with strangers if you don’t think they’ll be productive. You can prioritize the things that matter to you and make your life richer.

125117172964

Reposted by Gabriel Rocklin

Bryan Bryson @thebrysonlab.bsky.social · 05/11/2025

Excited to share our newest manuscript on antigen discovery and vaccine design for tuberculosis lead by Owen Leddy (everyone hiring new faculty in a few years, remember this name!) www.science.org/doi/10.1126/... #TBsky

science.org

Immunopeptidomics can inform the design of mRNA vaccines for the delivery of Mycobacterium tuberculosis MHC class II antigens

MHC class II immunopeptidomics in human phagocytes infected with Mycobacterium tuberculosis enables vaccine immunogen design and optimization.

1217

Gabriel Rocklin @grocklin.bsky.social · 24/11/2025

I don't know yet- clearly something has to do with strength of intermolecular interactions (e.g. the Arg effect) but also I bet something has to do with refolding rates. Although surprisingly we don't see much correlation with topology (% alpha vs % beta) so maybe not folding rates... ???

010

Gabriel Rocklin @grocklin.bsky.social · 21/11/2025

That Savas lab looks cool, but this was actually done in collaboration with Jeff Savas's lab at Northwestern, who is not on bsky that I can find :) thanks for the highlight!

030

I hope this is useful and we're excited to use this approach to explore more protein stresses- shelf life, vortexing, freeze-thaw, etc. I also love this paper because it's our lab's first time using quantitative proteomics! Congrats Cydney et al!!

090

Global Analysis of Aggregation Determinants in Small Protein Domains

All data are available on Zenodo and we'd love to see what you can do with it! Cydney also made a nice notebook to run the predictive model (including DMS scan) in Colab! Preprint: biorxiv.org/content/10.1... Colab: colab.research.google.com/drive/1KNWvG... Data: docs.google.com/forms/d/e/1F...

biorxiv.org

Protein aggregation is an obstacle for engineering effective recombinant proteins for biotechnology and therapeutic applications. Predicting protein aggregation propensity remains challenging due to t...

1103

There’s much more in the paper, including the surprising observation that higher folding stability correlates with aggregation even controlling for hydrophobicity (?!) Stabilities measured by Kotaro Tsuboyama!

260

Existing aggregation predictors show only modest correlations with our results, but fine tuning SaProt does well on our (carefully split) test set!

130

Arginine is strongly aggregation promoting! This was known in literature but not to me.

250

Hydrophobicity and isoelectric point (pI) are correlated with temp- and pH-propensity, but don’t tell the whole story. Other factors also influenced aggregation.

120

Really cool result: changes in soluble abundance from having all protein domains mixed together are nicely correlated with similar experiments on single, purified domains (also at 10 mg/mL).

140

Overall, 25-50% of the total protein became insoluble, but the proteomics revealed which proteins aggregated (became insoluble) more and which aggregated less.