Max Fürst
@maxfus.bsky.social
Asst. Prof. Uni Groningen 🇳🇱
Comp & Exp Biochemist, Protein Engineer, 'Would-be designer' (F. Arnold) | SynBio | HT Screens & Selections | Nucleic Acid Enzymes | Biocatalysis | Rstats & Datavis
https://www.fuerstlab.com
https://orcid.org/0000-0001-7720-9
Comp & Exp Biochemist, Protein Engineer, 'Would-be designer' (F. Arnold) | SynBio | HT Screens & Selections | Nucleic Acid Enzymes | Biocatalysis | Rstats & Datavis
https://www.fuerstlab.com
https://orcid.org/0000-0001-7720-9
Pinned
Max Fürst
@maxfus.bsky.social
· Dec 16
New preprint🚨
Imagine (re)designing a protein via inverse folding. AF2 predicts the designed sequence to a structure with pLDDT 94 & you get 1.8 Å RMSD to the input. Perfect design?
What if I told u that the structure has 4 solvent-exposed Trp and 3 Pro where a Gly should be?
Why to be wary🧵👇
Imagine (re)designing a protein via inverse folding. AF2 predicts the designed sequence to a structure with pLDDT 94 & you get 1.8 Å RMSD to the input. Perfect design?
What if I told u that the structure has 4 solvent-exposed Trp and 3 Pro where a Gly should be?
Why to be wary🧵👇
Reposted by Max Fürst
Want to get the data out of a PDF figure? As in, the actual data – not a rough trace-along-the-lines version?
I made an app you might like: adamkucharski.github.io/pdf2plot/
It all started a few years ago... 🧵
I made an app you might like: adamkucharski.github.io/pdf2plot/
It all started a few years ago... 🧵
January 13, 2026 at 9:12 PM
Want to get the data out of a PDF figure? As in, the actual data – not a rough trace-along-the-lines version?
I made an app you might like: adamkucharski.github.io/pdf2plot/
It all started a few years ago... 🧵
I made an app you might like: adamkucharski.github.io/pdf2plot/
It all started a few years ago... 🧵
with MD, with AI MDemulators, with AF hacks, ...
January 13, 2026 at 6:23 AM
with MD, with AI MDemulators, with AF hacks, ...
Thanks for sharing. Not surprised but good to see data also for antibodies that squares really well with our analysis on protein design more generally
bsky.app/profile/maxf...
bsky.app/profile/maxf...
New preprint🚨
Imagine (re)designing a protein via inverse folding. AF2 predicts the designed sequence to a structure with pLDDT 94 & you get 1.8 Å RMSD to the input. Perfect design?
What if I told u that the structure has 4 solvent-exposed Trp and 3 Pro where a Gly should be?
Why to be wary🧵👇
Imagine (re)designing a protein via inverse folding. AF2 predicts the designed sequence to a structure with pLDDT 94 & you get 1.8 Å RMSD to the input. Perfect design?
What if I told u that the structure has 4 solvent-exposed Trp and 3 Pro where a Gly should be?
Why to be wary🧵👇
January 10, 2026 at 7:34 PM
Thanks for sharing. Not surprised but good to see data also for antibodies that squares really well with our analysis on protein design more generally
bsky.app/profile/maxf...
bsky.app/profile/maxf...
The only format worse than pdb is cif
My hot take is that CIF and PDB formats are totally ill-equipped for adequately storing metadata describing how structures get generated and processed, and this problem will get worse as AI agents imbued with structural bioinformatics skills pass these files back and forth
January 9, 2026 at 6:33 PM
The only format worse than pdb is cif
Industrial computational protein engineering position in the south of the Netherlands
synsilico.com/storage/app/...
synsilico.com/storage/app/...
synsilico.com
January 9, 2026 at 7:19 AM
Industrial computational protein engineering position in the south of the Netherlands
synsilico.com/storage/app/...
synsilico.com/storage/app/...
Yea this is not the same of course, but it just popped up in my feed and was close enough in topic ;)
Still, there might be something there. Sample unbiased random flex and score with IF?
Still, there might be something there. Sample unbiased random flex and score with IF?
January 8, 2026 at 7:59 PM
Yea this is not the same of course, but it just popped up in my feed and was close enough in topic ;)
Still, there might be something there. Sample unbiased random flex and score with IF?
Still, there might be something there. Sample unbiased random flex and score with IF?
Reposted by Max Fürst
Can we design mutations that bias proteins towards desired conformational states?
Today in @science.org, we introduce Conformational Biasing (CB), a simple and scalable computational method that uses contrastive scoring by inverse folding models to identify conformation-biasing mutations.
Today in @science.org, we introduce Conformational Biasing (CB), a simple and scalable computational method that uses contrastive scoring by inverse folding models to identify conformation-biasing mutations.
Computational design of conformation-biasing mutations to alter protein functions
Conformational biasing (CB) is a rapid and streamlined computational method that uses contrastive scoring by inverse folding models to predict protein variants biased toward desired conformational sta...
www.science.org
January 8, 2026 at 7:08 PM
Can we design mutations that bias proteins towards desired conformational states?
Today in @science.org, we introduce Conformational Biasing (CB), a simple and scalable computational method that uses contrastive scoring by inverse folding models to identify conformation-biasing mutations.
Today in @science.org, we introduce Conformational Biasing (CB), a simple and scalable computational method that uses contrastive scoring by inverse folding models to identify conformation-biasing mutations.
Inverse folding much more promising
bsky.app/profile/alic...
bsky.app/profile/alic...
Video introduction to our new “Conformational Biasing” method for computational design of mutations that bias proteins towards desired conformational states
CB part starts at 14:55
Thanks to Peter Cavanagh and Andrew Xue – amazing graduate students who co-led this work
CB part starts at 14:55
Thanks to Peter Cavanagh and Andrew Xue – amazing graduate students who co-led this work
Alice Ting Rosettacon keynote 2025
YouTube video by Alice Ting
www.youtube.com
January 8, 2026 at 7:53 PM
Inverse folding much more promising
bsky.app/profile/alic...
bsky.app/profile/alic...
Personally skeptical of a paper making broad claims about such methods capturing mutation effects when their data is a single mutation in a single enzyme where effect is extensively studied & captured in many training data structures. AI
structure/dynamics prediction tools are very poor VAEs imo
structure/dynamics prediction tools are very poor VAEs imo
January 8, 2026 at 6:34 PM
Personally skeptical of a paper making broad claims about such methods capturing mutation effects when their data is a single mutation in a single enzyme where effect is extensively studied & captured in many training data structures. AI
structure/dynamics prediction tools are very poor VAEs imo
structure/dynamics prediction tools are very poor VAEs imo
Reposted by Max Fürst
Today I made a small but momentous start to work in 2026 by changing a single number.
I renamed the file “Papers to write and submit in 2025” to “Papers to write and submit in 2026”.
Stay tuned for more file updates on 1st January 2027.
I renamed the file “Papers to write and submit in 2025” to “Papers to write and submit in 2026”.
Stay tuned for more file updates on 1st January 2027.
January 1, 2026 at 4:26 PM
Today I made a small but momentous start to work in 2026 by changing a single number.
I renamed the file “Papers to write and submit in 2025” to “Papers to write and submit in 2026”.
Stay tuned for more file updates on 1st January 2027.
I renamed the file “Papers to write and submit in 2025” to “Papers to write and submit in 2026”.
Stay tuned for more file updates on 1st January 2027.
Reposted by Max Fürst
1/ Check out our newest paper where we ask: How fast can we experimentally discover binders from scratch?
And we mean scratch: a blinded study.
TLDR: 26 days. And the binders work…and led to new cancer biology.
We’re coming for you AI….
chemrxiv.org/engage/chemr...
And we mean scratch: a blinded study.
TLDR: 26 days. And the binders work…and led to new cancer biology.
We’re coming for you AI….
chemrxiv.org/engage/chemr...
December 26, 2025 at 6:24 PM
1/ Check out our newest paper where we ask: How fast can we experimentally discover binders from scratch?
And we mean scratch: a blinded study.
TLDR: 26 days. And the binders work…and led to new cancer biology.
We’re coming for you AI….
chemrxiv.org/engage/chemr...
And we mean scratch: a blinded study.
TLDR: 26 days. And the binders work…and led to new cancer biology.
We’re coming for you AI….
chemrxiv.org/engage/chemr...
Caught between excitement that "one of us" appears on a major podcast and cringe for parts of the discussion (mainly the self promotion, sadly a hallmark of otherwise enjoyable #hardfork)
Where Is All the A.I.-Driven Scientific Progress?
open.spotify.com
December 30, 2025 at 3:20 PM
Caught between excitement that "one of us" appears on a major podcast and cringe for parts of the discussion (mainly the self promotion, sadly a hallmark of otherwise enjoyable #hardfork)
Fantastic! Had 3 Qs after thread & think I understood from paper myself
1 were the 500 ligands v similar? - No, quite diverse
2 did they do "fair" real-life docking when comparing - i think so
3 is one limitation that all results are for one receptor & might not generalize? - probably
Would u agree?
1 were the 500 ligands v similar? - No, quite diverse
2 did they do "fair" real-life docking when comparing - i think so
3 is one limitation that all results are for one receptor & might not generalize? - probably
Would u agree?
December 30, 2025 at 10:28 AM
Fantastic! Had 3 Qs after thread & think I understood from paper myself
1 were the 500 ligands v similar? - No, quite diverse
2 did they do "fair" real-life docking when comparing - i think so
3 is one limitation that all results are for one receptor & might not generalize? - probably
Would u agree?
1 were the 500 ligands v similar? - No, quite diverse
2 did they do "fair" real-life docking when comparing - i think so
3 is one limitation that all results are for one receptor & might not generalize? - probably
Would u agree?
Reposted by Max Fürst
I'm really excited to break up the holiday relaxation time with a new preprint that benchmarks AlphaFold3 (AF3)/“co-folding” methods with 2 new stringent performance tests.
Thread below - but first some links:
A longer take:
fraserlab.com/2025/12/29/k...
Preprint:
www.biorxiv.org/content/10.6...
Thread below - but first some links:
A longer take:
fraserlab.com/2025/12/29/k...
Preprint:
www.biorxiv.org/content/10.6...
Know when to co-fold'em
This is the official web page for the James Fraser Lab at UCSF.
fraserlab.com
December 29, 2025 at 10:25 PM
I'm really excited to break up the holiday relaxation time with a new preprint that benchmarks AlphaFold3 (AF3)/“co-folding” methods with 2 new stringent performance tests.
Thread below - but first some links:
A longer take:
fraserlab.com/2025/12/29/k...
Preprint:
www.biorxiv.org/content/10.6...
Thread below - but first some links:
A longer take:
fraserlab.com/2025/12/29/k...
Preprint:
www.biorxiv.org/content/10.6...
Reposted by Max Fürst
So 2025 turned out to be a big year for de novo antibody design! Here are thoughts and predictions on the state of de novo antibody design heading into 2026 🧵
December 26, 2025 at 11:49 PM
So 2025 turned out to be a big year for de novo antibody design! Here are thoughts and predictions on the state of de novo antibody design heading into 2026 🧵
Reposted by Max Fürst
"How can funders avoid crossing the Szilard point?"
The Szilard point is "the threshold at which the total cost of competing for a grant equals (or surpasses) the value of the available funding."
The Szilard point is "the threshold at which the total cost of competing for a grant equals (or surpasses) the value of the available funding."
Point of no returns: researchers are crossing a threshold in the fight for funding
With so little money to go round, the costs of competing for grants can exceed what the grants are worth. When that happens, nobody wins.
www.nature.com
December 19, 2025 at 7:57 AM
"How can funders avoid crossing the Szilard point?"
The Szilard point is "the threshold at which the total cost of competing for a grant equals (or surpasses) the value of the available funding."
The Szilard point is "the threshold at which the total cost of competing for a grant equals (or surpasses) the value of the available funding."
Nice first step indeed. But were the ribosomal proteins ever the bottleneck? I'd expect rRNA and tRNAs first and can't rebuild without all the RNA modification enzymes. But probably IVTT anyway usually stops due to ATP depletion, right?
December 18, 2025 at 2:09 PM
Nice first step indeed. But were the ribosomal proteins ever the bottleneck? I'd expect rRNA and tRNAs first and can't rebuild without all the RNA modification enzymes. But probably IVTT anyway usually stops due to ATP depletion, right?
Precisely. And that's what we want to encourage: implement custom scoring suitable for your goal to try to mitigate the shortcoming of refolding metrics. Or, for novices who may not be able to do much in this regard, at least be skeptical of these metrics
December 18, 2025 at 2:03 PM
Precisely. And that's what we want to encourage: implement custom scoring suitable for your goal to try to mitigate the shortcoming of refolding metrics. Or, for novices who may not be able to do much in this regard, at least be skeptical of these metrics
Well many sequences require chaperones or only get structured in complex with something, but this is not so easy to compile data. Synthetic nonsense would probably be easiest but whether this can be well implemented in training without compromising positive performance is a question.
December 18, 2025 at 12:17 PM
Well many sequences require chaperones or only get structured in complex with something, but this is not so easy to compile data. Synthetic nonsense would probably be easiest but whether this can be well implemented in training without compromising positive performance is a question.
Ohh nice, hadn't seen that then! Might have reconsidered terminology in the title if I had.. 😅
December 18, 2025 at 11:24 AM
Ohh nice, hadn't seen that then! Might have reconsidered terminology in the title if I had.. 😅
I'd say so - AF2 just hasn't ever seen any negative data. We discuss this a bit in the paper although refrain from speculating to widely.
December 18, 2025 at 9:59 AM
I'd say so - AF2 just hasn't ever seen any negative data. We discuss this a bit in the paper although refrain from speculating to widely.
20/19
Or, you know, if bioRxiv is down / extremely slow again, download the pdf here:
www.fuerstlab.com/uploads/2025...
Or, you know, if bioRxiv is down / extremely slow again, download the pdf here:
www.fuerstlab.com/uploads/2025...
www.fuerstlab.com
December 18, 2025 at 9:57 AM
20/19
Or, you know, if bioRxiv is down / extremely slow again, download the pdf here:
www.fuerstlab.com/uploads/2025...
Or, you know, if bioRxiv is down / extremely slow again, download the pdf here:
www.fuerstlab.com/uploads/2025...
MSA struc as temp to ss I'd expect similar to MSA but maybe worth a shot. Energy based I guess will not fold nonsense, thus hard to implement in our tests except maybe in very low mutation regime; could test that in pMPNN background. Will think, thanks!
December 16, 2025 at 7:53 PM
MSA struc as temp to ss I'd expect similar to MSA but maybe worth a shot. Energy based I guess will not fold nonsense, thus hard to implement in our tests except maybe in very low mutation regime; could test that in pMPNN background. Will think, thanks!
Thanks to Seva and Kerlen for their tireless effort to get the data as watertight as possible.
Here is the preprint link again:
www.biorxiv.org/content/10.6...
19/19
Here is the preprint link again:
www.biorxiv.org/content/10.6...
19/19
Limitations of the refolding pipeline for de novo protein design
With the emergence of powerful deep learning-based tools, computational protein design has become a widely accessible technique. Nowadays, it is possible to perform both sequence and structure design ...
www.biorxiv.org
December 16, 2025 at 3:32 PM
Thanks to Seva and Kerlen for their tireless effort to get the data as watertight as possible.
Here is the preprint link again:
www.biorxiv.org/content/10.6...
19/19
Here is the preprint link again:
www.biorxiv.org/content/10.6...
19/19