Krishna Shrinivas
@shrinivaslab.bsky.social
Asst. prof at Northwestern ChBE
Interested in how molecules and processes are organized and regulated in living cells | physics, math, engineering, and computation (mostly) for biology
shrinivaslab.com
Interested in how molecules and processes are organized and regulated in living cells | physics, math, engineering, and computation (mostly) for biology
shrinivaslab.com
Congrats Ben, Hue Sun, and all authors!
October 22, 2025 at 6:33 PM
Congrats Ben, Hue Sun, and all authors!
Thanks karthik :)
October 11, 2025 at 12:03 AM
Thanks karthik :)
Our framework:
We bridge machine learning & statistical physics to directly invert molecular simulations to design IDPS and engineer examples that:
🌀 form loops & linkers with tuned flexibility
⚡ sense salt, temperature, or phosphorylation stimuli
🤝 bind disordered targets like FUS or Whi3
We bridge machine learning & statistical physics to directly invert molecular simulations to design IDPS and engineer examples that:
🌀 form loops & linkers with tuned flexibility
⚡ sense salt, temperature, or phosphorylation stimuli
🤝 bind disordered targets like FUS or Whi3
October 10, 2025 at 6:16 PM
Our framework:
We bridge machine learning & statistical physics to directly invert molecular simulations to design IDPS and engineer examples that:
🌀 form loops & linkers with tuned flexibility
⚡ sense salt, temperature, or phosphorylation stimuli
🤝 bind disordered targets like FUS or Whi3
We bridge machine learning & statistical physics to directly invert molecular simulations to design IDPS and engineer examples that:
🌀 form loops & linkers with tuned flexibility
⚡ sense salt, temperature, or phosphorylation stimuli
🤝 bind disordered targets like FUS or Whi3
The problem:
AI tools like AlphaFold & ProteinMPNN accelerate design of stable protein folds by inverting the sequence-structure map.
But IDPs don't have 1 shape - they occupy a huge ensemble of shapes. Physics simulations are good models to generate ensembles but hard to design/invert over!
AI tools like AlphaFold & ProteinMPNN accelerate design of stable protein folds by inverting the sequence-structure map.
But IDPs don't have 1 shape - they occupy a huge ensemble of shapes. Physics simulations are good models to generate ensembles but hard to design/invert over!
October 10, 2025 at 6:16 PM
The problem:
AI tools like AlphaFold & ProteinMPNN accelerate design of stable protein folds by inverting the sequence-structure map.
But IDPs don't have 1 shape - they occupy a huge ensemble of shapes. Physics simulations are good models to generate ensembles but hard to design/invert over!
AI tools like AlphaFold & ProteinMPNN accelerate design of stable protein folds by inverting the sequence-structure map.
But IDPs don't have 1 shape - they occupy a huge ensemble of shapes. Physics simulations are good models to generate ensembles but hard to design/invert over!
Many congrats Alex! Your labs research has been a pleasure to read (and try code openly). Hope you are celebrating 🍾
October 3, 2025 at 12:52 AM
Many congrats Alex! Your labs research has been a pleasure to read (and try code openly). Hope you are celebrating 🍾
Led by the amazing Aidan Zentner, with contribs from Ethan Halingstad, and in collab with Cameron Chalk, Michael Brenner, @amurugan.bsky.social, and Erik Winfree.
For a more fun overview, see Erik's version of the abstract www.dna.caltech.edu/DNAresearch_... :) (2/2)
For a more fun overview, see Erik's version of the abstract www.dna.caltech.edu/DNAresearch_... :) (2/2)
September 22, 2025 at 9:38 PM
Led by the amazing Aidan Zentner, with contribs from Ethan Halingstad, and in collab with Cameron Chalk, Michael Brenner, @amurugan.bsky.social, and Erik Winfree.
For a more fun overview, see Erik's version of the abstract www.dna.caltech.edu/DNAresearch_... :) (2/2)
For a more fun overview, see Erik's version of the abstract www.dna.caltech.edu/DNAresearch_... :) (2/2)
congrats Amy!
September 13, 2025 at 2:07 AM
congrats Amy!
This project began at the @mblscience.bsky.social Physiology course 🌊 and grew into a FUN collaboration over many years across Northwestern and Duke - thanks all for the support and more coverage below!
📰 Coverage:
www.mccormick.northwestern.edu/news/article...
www.mbl.edu/news/physiol...
📰 Coverage:
www.mccormick.northwestern.edu/news/article...
www.mbl.edu/news/physiol...
Cell Feature Implicated in Cancer Forms Differently than Previously Thought
A team with Professor Krishna Shrinivas discovered how a landmark found in the nucleus of cells in many organ systems forms through a different mechanism than the well-established view, a revelation t...
www.mccormick.northwestern.edu
September 4, 2025 at 2:44 AM
This project began at the @mblscience.bsky.social Physiology course 🌊 and grew into a FUN collaboration over many years across Northwestern and Duke - thanks all for the support and more coverage below!
📰 Coverage:
www.mccormick.northwestern.edu/news/article...
www.mbl.edu/news/physiol...
📰 Coverage:
www.mccormick.northwestern.edu/news/article...
www.mbl.edu/news/physiol...
Extra curiosities 🔍
• Across tissues & species, stoichiometries of NONO/FUS are conserved, hinting at evolutionary tuning.
• Simulations by Mary Skillicorn in the lab also suggest important roles for co-transcriptional nucleation of paraspeckles for tuning paraspeckle size/number.
• Across tissues & species, stoichiometries of NONO/FUS are conserved, hinting at evolutionary tuning.
• Simulations by Mary Skillicorn in the lab also suggest important roles for co-transcriptional nucleation of paraspeckles for tuning paraspeckle size/number.
September 4, 2025 at 2:44 AM
Extra curiosities 🔍
• Across tissues & species, stoichiometries of NONO/FUS are conserved, hinting at evolutionary tuning.
• Simulations by Mary Skillicorn in the lab also suggest important roles for co-transcriptional nucleation of paraspeckles for tuning paraspeckle size/number.
• Across tissues & species, stoichiometries of NONO/FUS are conserved, hinting at evolutionary tuning.
• Simulations by Mary Skillicorn in the lab also suggest important roles for co-transcriptional nucleation of paraspeckles for tuning paraspeckle size/number.
Another surprise: core & shell proteins don’t mix well (they’re immiscible, like oil & water).
Putting these observations together in simulations suggests 🖥️⚛️: competition for RNA + immiscibility naturally push proteins to form different layers, even if they individually like the same parts of RNA.
Putting these observations together in simulations suggests 🖥️⚛️: competition for RNA + immiscibility naturally push proteins to form different layers, even if they individually like the same parts of RNA.
September 4, 2025 at 2:44 AM
Another surprise: core & shell proteins don’t mix well (they’re immiscible, like oil & water).
Putting these observations together in simulations suggests 🖥️⚛️: competition for RNA + immiscibility naturally push proteins to form different layers, even if they individually like the same parts of RNA.
Putting these observations together in simulations suggests 🖥️⚛️: competition for RNA + immiscibility naturally push proteins to form different layers, even if they individually like the same parts of RNA.
We combined in vitro assays of binding and condensation with bioinformatics to ask which parts of NEAT1 each protein preferred binding to.
Surprise: core proteins (FUS, NONO) actually prefer the same shell RNA regions as the shell protein TDP-43! Everyone crowds into the same RNA zones. 🌀
Surprise: core proteins (FUS, NONO) actually prefer the same shell RNA regions as the shell protein TDP-43! Everyone crowds into the same RNA zones. 🌀
September 4, 2025 at 2:44 AM
We combined in vitro assays of binding and condensation with bioinformatics to ask which parts of NEAT1 each protein preferred binding to.
Surprise: core proteins (FUS, NONO) actually prefer the same shell RNA regions as the shell protein TDP-43! Everyone crowds into the same RNA zones. 🌀
Surprise: core proteins (FUS, NONO) actually prefer the same shell RNA regions as the shell protein TDP-43! Everyone crowds into the same RNA zones. 🌀
Prevailing model suggests:
→ “Core” proteins bind the middle of the NEAT1 RNA scaffold
→ “Shell” proteins bind the RNA ends
This selective binding could, in principle, assemble layers - but has not been explicitly tested. So we set out to do this!
→ “Core” proteins bind the middle of the NEAT1 RNA scaffold
→ “Shell” proteins bind the RNA ends
This selective binding could, in principle, assemble layers - but has not been explicitly tested. So we set out to do this!
September 4, 2025 at 2:44 AM
Prevailing model suggests:
→ “Core” proteins bind the middle of the NEAT1 RNA scaffold
→ “Shell” proteins bind the RNA ends
This selective binding could, in principle, assemble layers - but has not been explicitly tested. So we set out to do this!
→ “Core” proteins bind the middle of the NEAT1 RNA scaffold
→ “Shell” proteins bind the RNA ends
This selective binding could, in principle, assemble layers - but has not been explicitly tested. So we set out to do this!
We use paraspeckles as a model to study this question.
Paraspeckles are built around a non-coding RNA NEAT1 whose middle regions are in the core and 5’/3’ ends on the shell layer - each layer also recruiting different proteins. (2/6)
Paraspeckles are built around a non-coding RNA NEAT1 whose middle regions are in the core and 5’/3’ ends on the shell layer - each layer also recruiting different proteins. (2/6)
September 4, 2025 at 2:44 AM
We use paraspeckles as a model to study this question.
Paraspeckles are built around a non-coding RNA NEAT1 whose middle regions are in the core and 5’/3’ ends on the shell layer - each layer also recruiting different proteins. (2/6)
Paraspeckles are built around a non-coding RNA NEAT1 whose middle regions are in the core and 5’/3’ ends on the shell layer - each layer also recruiting different proteins. (2/6)