Andrew Savinov

Thrilled to announce that the Savinov lab is now live at UT Austin @utaustin.bsky.social! Our lab employs massively parallel experimental and AI methods to discover and design protein fragments as universal regulators of protein interactions in health and disease. www.savinovlab.com (1/n)

Andrew Savinov reposted

Excited to announce that the Rikki Garner lab has opened its doors in beautiful Austin, TX! Our interdisciplinary lab @utaustin.bsky.social MBS investigates tissue fluidity as a universal physical regulator of multicellular patterning in health and disease. www.rikkigarnerlab.org 1/n

savinovlab.comPeople — The Savinov Lab at UT Austin

We are also excited to hire a highly motivated and organized individual to push forward exciting work in human cell culture and act as lab manager tinyurl.com/yc5c5p2f. Keep an eye on other openings at www.savinovlab.com/people and feel free to reach out directly! (4/n)

We are passionate about unlocking the hidden biology & powerful applications of protein fragments – enabling us to study and control cellular protein interactions at unprecedented scale. If that sounds exciting, please follow our work, join us, or reach out to collaborate! (5/n)

Come join us in pushing the boundaries of this exciting new field! We are looking for curious and passionate scientists at all levels. We are an interdisciplinary lab interested in backgrounds from cell and molecular biology to biophysics, engineering, and computer science. (2/n)

ils.utexas.eduGraduate Study in Life SciencesInterdisciplinary life sciences programs at The University of Texas at Austin

We are looking for postdocs and grad students! If you are interested in pursuing a Ph.D. in the lab, we are recruiting through the Interdisciplinary Life Sciences graduate program ils.utexas.edu & have growing connections w/ Computer Science & Physics among others. (3/n)

Interested in genetically encodable inhibitors of your favorite biomolecular condensate? Excited to announce our latest work, w/ Jibin Sadasivan, Gene-Wei Li, & Lindsay Case, on protein fragments as generalizable regulators of phase separation. (1/n) www.biorxiv.org/content/10.6...

This work has been a fantastic collaboration w/ co-first Jibin Sadasivan + Jack Rubien and Kyle White & initially arose from a conversation w/ co-corresponding Lindsay Case at the Structure-Function Supergroup @mit.edu! As always, great to work w/ Gene-Wei Li. (17/n)

Thanks to NIH and @hhmi-science.bsky.social for supporting this work! (18/n) @mit.edu

Overall, our results establish AI-driven protein fragment discovery as a generalizable strategy to dissect and control the molecular interactions that govern biomolecular condensates – with many fundamental and therapeutic applications. (15/n)

Our results further highlight that protein fragments are universal regulators of protein interactions, from bacteria to human. Protein fragments hold immense promise as tools to study and control cellular protein interactions across species and systems. (16/n)

We validated this novel driving interaction by deleting the kinase domain and showing loss of condensate formation. (13/n)

We further demonstrated that the inhibitory fragment of FAK robustly titrates condensate formation – and is able to abrogate FAK condensate formation in living mammalian cells! We are excited about the implications for novel cancer therapeutics. (14/n)

Our systematic approach therefore successfully identified NTF2L domain dimerization as essential for condensate formation while simultaneously providing novel peptide inhibitors to control these interactions! (11/n)

In the case of FAK, a cancer driver protein that forms condensates at focal adhesions, we discovered both condensate-inhibiting and -enhancing fragments. The inhibitory fragment uncovered a novel interaction between the FERM and kinase domains driving phase separation! (12/n)

Notably, these results are obtained with no additional modifications required from the linear, genetically encodable peptide sequence! (9/n)

For example, we discover several fragments of G3BP1 that strongly inhibit condensate formation. FragFold predicts that these fragments bind in a native-like mode mimicking NTF2L dimerization, which is essential for phase separation and stress granule formation. (10/n)

We applied this approach across a range of human & viral proteins w/ important roles in health & disease: G3BP1, SARS-CoV-2 nucleocapsid, TDP-43, and focal adhesion kinase (FAK). Computationally screening 2,235 fragments, we selected 18 candidates for further investigation. (7/n)

Overall, we observed a 50% success rate (9/18 designs) in discovering condensate-controlling protein fragments, experimentally testing just 3-5 candidates per protein! And for each condensate-forming protein, the success rate was at least 40%. (8/n)

Given that biomolecular condensates are driven by multivalent interactions, we reasoned that fragments of condensate-forming proteins should provide a universal approach to control biomolecular condensates and study the molecular interactions driving phase separation. (5/n)

We therefore applied FragFold to predict how all possible fragments of diverse phase-separating proteins bind to their parental proteins, modulating condensate formation. We then selected a small set of AI-discovered protein fragments for experimental testing. (6/n)

pnas.orgMapping functional regions of essential bacterial proteins with dominant-negative protein fragments | PNASMassively parallel measurements of dominant-negative inhibition by protein fragments have been used to map protein interaction sites and discover p...

We have previously shown that small fragments of larger proteins are broadly functional as inhibitors of protein interactions, over >10,000 fragments of highly diverse proteins, directly in living cells. (www.pnas.org/doi/10.1073/...) (3/n)

In subsequent work, we developed the FragFold AI method to discover functional protein fragments, showing we can predict these fragments and their precise molecular binding modes on a massively parallel scale! (www.pnas.org/doi/10.1073/...) (4/n)

Biomolecular condensates are a major driver of cellular organization, with important physiological functions. However, we have lacked a predictable and systematic approach to modulate the multivalent interactions underlying their formation. (2/n)

Andrew Savinov reposted

@katecavanaugh.bsky.social

5 Dec 2025

If you're at #cellbio2025, @atmolines.bsky.social Hernan Garcia and I invite you to attend our Minisymposium “Physical Cell Biology from Molecules to Organisms” for incredible talks on epigenetic mechanical memory, viscoelasticity, cortical flows, morphogenesis, tissue wetting, size scaling, et al!

Andrew Savinov reposted

Kate Cavanaugh

6 Dec 2025

🚨 Link up with me at #Cellbio2025 ! 🚨 Excited to present the majority of my postdoc work in the Physical Cell Biology from Molecules to Organisms Minisymposium. Thanks to the organizers for putting together such an exciting session 🤩

Rikki Garner@rikkigarner.bsky.social

Andrew Savinov reposted

Liam J. Russell

@liamjrussell99.bsky.social

6 Dec 2025

Super excited to share some of my work in such a great session at #cellbio2025!! If you’re at the meeting come on down, I’ll be talking about cell density and its influence on tissue fluidization during epithelial morphogenesis 🔬👨🏻‍🔬🪰

Rikki Garner@rikkigarner.bsky.social

Andrew Savinov reposted

5 Dec 2025

Looking forward to seeing everyone at #cellbio2025! Excited to share new work from my postdoc to measure, model, and control the interplay of tissue fluidity and multicellular patterning in a living organism. Come see me at my talk or poster (details below), or DM me if you want to meet up!

Thrilled to contribute to this exciting work on protein binder design together w/ @hannes-stark.bsky.social , Felix Faltings, Regina Barzilay, Tommi Jaakkola, and co. We applied BoltzGen to design novel antimicrobial peptides targeting DNA gyrase based on inhibitory protein fragments. (1/n)

Hannes Stark@hannes-stark.bsky.social

Excited to release BoltzGen which brings SOTA folding performance to binder design! The best part of this project is collaborating with a broad network of leading wetlabs that test BoltzGen at an unprecedented scale, showing success on many novel targets and pushing the model to its limits!

Almost all of these de novo anti-gyrase peptides were so strongly toxic that they bacteria expressing them were totally eliminated from the population – similar to the inhibitory fragments we previously discovered! We're excited about these peptides as novel antimicrobials. (4/n)

As always, huge thanks to Gene-Wei Li with whom it was fantastic to do this work, & of course @hannes-stark.bsky.social, Felix Faltings, Regina Barzilay, and Tommi Jaakkola for the collaboration. @MITBiology @hhmi.org (5/n)

Hannes Stark@hannes-stark.bsky.social

pnas.orgHigh-throughput discovery of inhibitory protein fragments with AlphaFold | PNASPeptides can bind to specific sites on larger proteins and thereby function as inhibitors and regulatory elements. Peptide fragments of larger prot...

We previously showed protein fragments are generalizable inhibitors of protein interactions (www.pnas.org/doi/10.1073/...). Here, show that target sites discovered by protein fragment scanning can be attacked by alternative modalities – in particular de novo peptide binders! (2/n)

We tested ~1800 designs experimentally in living cells alongside fragments tiling across DNA gyrase. About 20% inhibited bacterial growth, and 5.5% were specific to the designed binding mode, which we assayed by almost 1800 mutants breaking the peptide binding interfaces! (3/n)

Andrew Savinov reposted

23 Oct 2025

Happy to share that this work is now published in @biophysj.bsky.social! doi.org/10.1016/j.bp...

Rikki Garner@rikkigarner.bsky.social

In developing embryos, cells move a lot! Plenty of that movement is random. Is random cell mixing a feature or bug for tissue patterning? Turns out, it’s both! Excited to share the 1st preprint from my postdoc w/ Sean Megason @seanemcgeary.bsky.social and Allon Klein. 1/20 doi.org/10.1101/2025...

Andrew Savinov reposted

10 Sept 2025

So honored to be speaking today at the Stanford.Berkeley.UCSF Next Generation Faculty Symposium! Hope to see you there! (Zoom registration link below)

Polly Fordyce@pollyfordyce.bsky.social

Incredibly excited for this year's iteration of the Next Generation Faculty Symposium on 9/10/25! This is a fantastic opportunity to hear 10 minute talks describing fantastic cutting-edge science from researchers preparing to launch a faculty search. Please register (link below) & forward widely :)

Andrew Savinov reposted

2 Sept 2025

Friendly reminder that the deadline to submit an abstract for talk at #cellbio2025 is TOMORROW, Wednesday, September 3, 2025 at 9 PM ET! Please consider submitting to our minisymposia "Physical Cell Biology from Molecules to Organisms (more details in the repost)

Rikki Garner@rikkigarner.bsky.social

Excited to announce our minisymp. “Physical Cell Biology from Molecules to Organisms” #CellBio2025 w/ Hernan Garcia and ‪@atmolines.bsky.social‬! Have an interesting story on how physical cell properties are controlled, coordinated across scales, or contribute to development and disease? (1/3)

MIT Department of Chemistry

4 Aug 2025

Check out this exciting new minisymposium at ASCB!

Rikki Garner@rikkigarner.bsky.social

Andrew Savinov reposted

@mitchemistry.bsky.social

2 July 2025

Congratulations to Barbara Imperiali, the Class of 1922 Professor of Chemistry and Biology, who has been elected as a Fellow of the Royal Society, the United Kingdom’s national academy of sciences. chemistry.mit.edu/chemistry-ne...

Andrew Savinov reposted

Chrystal Starbird

@drstarbird.bsky.social

26 June 2025

The @berlowlab.bsky.social kicking off the next session at #PS39 If you missed his talk, Jake Simmons and Jamie Do of the Berlow lab will both be presenting posters at the afternoon session!

Andrew Savinov reposted

Debraj Ghose

@debrajghose.bsky.social

5 May 2025

We found that cells use emergent collective memory—arising from simple chemical reactions—to outperform physical limits of detecting chemical gradients. Curious how? 1/n 8 min talk: www.youtube.com/watch?v=A7XH... bioRxiv: www.biorxiv.org/content/10.1... #biophysics #cellbiology

youtube.comLocal collective memory from ratiometric signaling outperforms cellular gradient sensing limitsYouTube video by Debraj Shubham Ghose

@whiteheadinstitute.bsky.social

9 June 2025

Excited to be at the @bostonbacteria.bsky.social‬ Meeting this year! I'll be a panelist at the AI in biology breakout session on Tuesday 6/10 after lunch, in Hall A -- together with ‪@microyunha.bsky.social‬ and M.S. Suryateja Jammalamadaka. Come join us for what should be a great discussion!

Andrew Savinov reposted

Whitehead Institute

30 Apr 2025

Whitehead Institute Member Yukiko Yamashita, who studies how genetic information is passed from parents to offspring, has been elected to the National Academy of Sciences. Full story on our website: wi.mit.edu/news/yukiko-...

Andrew Savinov reposted

Arjun Raj

@arjunraj.bsky.social

28 Apr 2025

So excited to see this live! Awesome work from @giannatbusch.bsky.social on the concept of population-based synergy—using cell-to-cell heterogeneity to uncover new therapeutic vulnerabilities!

Gianna T. Busch@giannatbusch.bsky.social

Excited to share our new work on the distinct drug sensitivity profiles of heterogeneous therapy-resistant melanoma cells! @arjunraj.bsky.social www.biorxiv.org/content/10.1... (1/10)

biorxiv.orgHeterogeneous therapy-resistant cancer cells have distinct and exploitable drug sensitivity profilesResistance to targeted therapies is a significant clinical problem, but eliminating resistant cancer cells has proven difficult. One potential reason for this difficulty is heterogeneity in the resist...

Andrew Savinov reposted

Nick Polizzi

@nickpolizzi.bsky.social

28 Apr 2025

Super excited to share a new preprint from our lab on design of small-molecule binding proteins using neural networks! The paper has a bit of everything. A new graph neural network, new design algorithms, and experimental validation. www.biorxiv.org/content/10.1... 🧵🧪

biorxiv.orgZero-shot design of drug-binding proteins via neural selection-expansionComputational design of molecular recognition remains challenging despite advances in deep learning. The design of proteins that bind to small molecules has been particularly difficult because it requ...

Andrew Savinov reposted

Alex Bisson

@archaeon-alex.bsky.social

24 Apr 2025

I'm the biggest fan of Rikki. Elegant physics. Breathtaking biology. If you're looking for a seminar speaker...she's a fantastic communicator. Congrats for the new work, @rikkigarner.bsky.social!

Rikki Garner@rikkigarner.bsky.social

24 Apr 2025

Check out this extremely thought-provoking work on tissue fluidity, cell motility, and patterning from @rikkigarner.bsky.social! Certainly changes how I think about development...

Rikki Garner@rikkigarner.bsky.social

biology.mit.eduA sum of their partsAll biological function is dependent on how different proteins interact with each other. Protein-protein interactions facilitate everything from transcribing DNA and controlling cell division to highe...

11 Feb 2025

Also happy to share the @MITBiology news brief highlighting our @NIH-funded basic research which paves the way to systematically develop novel inhibitors and drugs, such as peptide-based antibiotics: biology.mit.edu/a-sum-of-the....

Andrew Savinov@biosavinov.bsky.social

Happy to announce that our paper on systematically discovering peptides to inhibit protein-protein interactions in living cells is out now @pnas.org: www.pnas.org/doi/10.1073/.... @MITBiology @hhmi.bsky.social See thread on the updated preprint for details: x.com/bioSavinov/s...

pnas.orgpnas.org

5 Feb 2025

pnas.orgpnas.org