Bryan Dickinson
@chembiobryan.bsky.social
chemical/synthetic biologist, Luddite trying to find better ways to make molecules that do important stuff, dad, @uchicago professor of chemistry
http://www.dickinsonlab.uchicago.edu/
http://www.dickinsonlab.uchicago.edu/
And it worked! Now we have binders that use a secondary hot spot, but that is LC3B selective. This demonstrates the system can target specific epitopes/regions within a single protein - precision at the molecular level! 15/n
November 19, 2025 at 6:10 PM
And it worked! Now we have binders that use a secondary hot spot, but that is LC3B selective. This demonstrates the system can target specific epitopes/regions within a single protein - precision at the molecular level! 15/n
So… we used PANCS-Spec-binders to force the selections to find binders that bind the next hottest hot spot on LC3B that IS NOT the LIR motif. This is an “Epitope-specific” selection. 14/n
November 19, 2025 at 6:10 PM
So… we used PANCS-Spec-binders to force the selections to find binders that bind the next hottest hot spot on LC3B that IS NOT the LIR motif. This is an “Epitope-specific” selection. 14/n
We figured out this is because the LC3B binders decided to bind at the “LIR motif”, which is shared between GABARAP and LC3B (side note, the LC3B binders were inherently selective – so interesting!). 13/n
November 19, 2025 at 6:10 PM
We figured out this is because the LC3B binders decided to bind at the “LIR motif”, which is shared between GABARAP and LC3B (side note, the LC3B binders were inherently selective – so interesting!). 13/n
Proof of Concept #2 - LC3B Region Targeting: As a second example, we developed binders specific to the LIR (LC3 Interacting Region) of LC3B. This stemmed from our initial PANCS-binder paper, where all our binders were incredibly selective…except for one. Our LC3B binder also bound GABARAP. =( 12/n
November 19, 2025 at 6:10 PM
Proof of Concept #2 - LC3B Region Targeting: As a second example, we developed binders specific to the LIR (LC3 Interacting Region) of LC3B. This stemmed from our initial PANCS-binder paper, where all our binders were incredibly selective…except for one. Our LC3B binder also bound GABARAP. =( 12/n
Swapping this since amino acid between HRas and KRas completely inverts the specificity! Not what we would have guessed, but that is the power of unbiased discovery. 11/n
November 19, 2025 at 6:10 PM
Swapping this since amino acid between HRas and KRas completely inverts the specificity! Not what we would have guessed, but that is the power of unbiased discovery. 11/n
How do out binders bind HRas selectively? Mapping the binding modes of our initial non-isoform selective was relatively easy. Alphafold predicted the site of binding, which we confirmed biochemically and by a really nice X-ray structure…9/n
November 19, 2025 at 6:10 PM
How do out binders bind HRas selectively? Mapping the binding modes of our initial non-isoform selective was relatively easy. Alphafold predicted the site of binding, which we confirmed biochemically and by a really nice X-ray structure…9/n
Proof of Concept #1 - RAS Specificity: In our initial PANCS-Binders paper, the binders we got that bound HRas also bind KRas (like N-LHY). This makes sense, since they are almost identical and share a key hotspot. Now, we can create binders like N-WYN that specifically bind HRAS! 8/n
November 19, 2025 at 6:10 PM
Proof of Concept #1 - RAS Specificity: In our initial PANCS-Binders paper, the binders we got that bound HRas also bind KRas (like N-LHY). This makes sense, since they are almost identical and share a key hotspot. Now, we can create binders like N-WYN that specifically bind HRAS! 8/n
How It Works: The system uses iterative rounds of simultaneous positive selection (bind your target) AND negative selection (avoid similar proteins). 6/n
November 19, 2025 at 6:10 PM
How It Works: The system uses iterative rounds of simultaneous positive selection (bind your target) AND negative selection (avoid similar proteins). 6/n
Why This Matters: Many proteins share "hot spots" - common binding surfaces that make specific targeting nearly impossible. This is especially problematic for protein families like RAS (crucial in cancer), where isoforms are >90% identical but have distinct biological roles. 5/n
November 19, 2025 at 6:10 PM
Why This Matters: Many proteins share "hot spots" - common binding surfaces that make specific targeting nearly impossible. This is especially problematic for protein families like RAS (crucial in cancer), where isoforms are >90% identical but have distinct biological roles. 5/n
The Problem: Protein binders are crucial research tools, but they often suffer from promiscuous binding - hitting multiple similar proteins when you only want ONE specific target. Or, binding your protein, but at the wrong site. Enter: PANCS-spec-Binders! 2/n
November 19, 2025 at 6:10 PM
The Problem: Protein binders are crucial research tools, but they often suffer from promiscuous binding - hitting multiple similar proteins when you only want ONE specific target. Or, binding your protein, but at the wrong site. Enter: PANCS-spec-Binders! 2/n
Why does this matter? We present a conceptual, experimental, and computational framework to understand selection success in binder discovery. We turn what is usually more of an art into a science. 9/12
October 15, 2025 at 5:19 PM
Why does this matter? We present a conceptual, experimental, and computational framework to understand selection success in binder discovery. We turn what is usually more of an art into a science. 9/12
Machine learning trained on our data – critically using contrastive learning with our vast non-binder datasets - can successfully predict binders, but needs >50% of the data to work well - showing you need large datasets to understand these vast and diverse landscapes. 8/12
October 15, 2025 at 5:19 PM
Machine learning trained on our data – critically using contrastive learning with our vast non-binder datasets - can successfully predict binders, but needs >50% of the data to work well - showing you need large datasets to understand these vast and diverse landscapes. 8/12
Each target has a unique "fitness landscape" with distinct binding clusters. These landscapes are highly reproducible - run the same selection twice, get the same answer in terms of fitness landscapes and clusters – even though the specific sequences will be different every time! 7/12
October 15, 2025 at 5:19 PM
Each target has a unique "fitness landscape" with distinct binding clusters. These landscapes are highly reproducible - run the same selection twice, get the same answer in terms of fitness landscapes and clusters – even though the specific sequences will be different every time! 7/12
The simplest randomization strategy (using all 20 amino acids equally) is nearly optimal! Fancy biased designs showed negligible improvements. Nature already optimized this. 6/12
October 15, 2025 at 5:19 PM
The simplest randomization strategy (using all 20 amino acids equally) is nearly optimal! Fancy biased designs showed negligible improvements. Nature already optimized this. 6/12
We found target identity matters WAY more than how you design your library. "Easy" targets need only 2-3 amino acids; "hard" targets need 6+ fixed position or have strict topographical requirements. 5/12
October 15, 2025 at 5:19 PM
We found target identity matters WAY more than how you design your library. "Easy" targets need only 2-3 amino acids; "hard" targets need 6+ fixed position or have strict topographical requirements. 5/12
Why are some targets “easy” and some “hard”? Minimal binding motifs of just 2-8 specified amino acids determine success. Fewer amino acids needed in a motif = a more common motif in a theoretical library = easier to find binders. It's surprisingly simple! 4/12
October 15, 2025 at 5:19 PM
Why are some targets “easy” and some “hard”? Minimal binding motifs of just 2-8 specified amino acids determine success. Fewer amino acids needed in a motif = a more common motif in a theoretical library = easier to find binders. It's surprisingly simple! 4/12
We created a framework to quantify “target difficulty” and found the likelihood of a random protein variant binding a target varies WILDLY - from 1 in 10,000 to over 1 in 10 billion! Most targets have a binding frequency around 1 in 100 million. 3/12
October 15, 2025 at 5:19 PM
We created a framework to quantify “target difficulty” and found the likelihood of a random protein variant binding a target varies WILDLY - from 1 in 10,000 to over 1 in 10 billion! Most targets have a binding frequency around 1 in 100 million. 3/12
To answer this, we combined our PANCS-Binder tech with advanced computational approaches to perform and analyze >1,300 selections testing how random protein variants bind to 96 different targets. 2/12
October 15, 2025 at 5:19 PM
To answer this, we combined our PANCS-Binder tech with advanced computational approaches to perform and analyze >1,300 selections testing how random protein variants bind to 96 different targets. 2/12
Thank you so much to the University of Kansas Department of Chemistry, Chembio program, COBRE program, and most importantly the trainees who invited me to take part in the 4th Annual Chemical Biology Symposium. Fantastic science and a wonderful way to spend a weekend. And congrats to awardees!!
October 11, 2025 at 10:12 PM
Thank you so much to the University of Kansas Department of Chemistry, Chembio program, COBRE program, and most importantly the trainees who invited me to take part in the 4th Annual Chemical Biology Symposium. Fantastic science and a wonderful way to spend a weekend. And congrats to awardees!!
Good balanced take on AI in protein design...hype and potential:
blog.genesmindsmachines.com/p/we-still-c...
blog.genesmindsmachines.com/p/we-still-c...
October 7, 2025 at 4:36 PM
Good balanced take on AI in protein design...hype and potential:
blog.genesmindsmachines.com/p/we-still-c...
blog.genesmindsmachines.com/p/we-still-c...
A successful NSF grant in 2025 deserves a nice dinner celebration. Thanks to this incredible crew of rockstars
September 23, 2025 at 1:10 AM
A successful NSF grant in 2025 deserves a nice dinner celebration. Thanks to this incredible crew of rockstars
So proud of our undergrad Josh Pixley defending his senior thesis @uchicagopme.bsky.social @uchichemistry.bsky.social today. So much energy and passion. I do hope our world capitalizes on the potential of Josh and all the other developing young scientists.
May 9, 2025 at 9:09 PM
So proud of our undergrad Josh Pixley defending his senior thesis @uchicagopme.bsky.social @uchichemistry.bsky.social today. So much energy and passion. I do hope our world capitalizes on the potential of Josh and all the other developing young scientists.
We are prepping for lean times with creative solutions. Came in today to find these 12 selections of billion+ variant libraries cooking, each run in 10 cent Costco water bottles
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
March 6, 2025 at 2:49 PM
We are prepping for lean times with creative solutions. Came in today to find these 12 selections of billion+ variant libraries cooking, each run in 10 cent Costco water bottles
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
When the world is infuriating and depressing … great time to hangout with your kickass chembio colleagues to vent, laugh, and celebrate science
February 21, 2025 at 3:18 AM
When the world is infuriating and depressing … great time to hangout with your kickass chembio colleagues to vent, laugh, and celebrate science