Kresten Lindorff-Larsen
LightNews LightNews
banner
lindorfflarsen.bsky.social
Kresten Lindorff-Larsen
@lindorfflarsen.bsky.social
5.5K followers 1.2K following 1.1K posts
Protein and coffee lover, father of two, professor of biophysics and sudo scientist at the Linderstrøm-Lang Centre for Protein Science, University of Copenhagen 🇩🇰
Posts Replies Media Videos
lindorfflarsen.bsky.social
@odedrechavi.bsky.social classic
November 4, 2025 at 8:33 PM
lindorfflarsen.bsky.social
I’m excited to be speaking at the AI in Science Summit 2025 in my home town of Copenhagen #AIS25 #AI4Science

I will be speaking about how AI and biophysics can come together to study protein dynamics
November 3, 2025 at 8:09 AM
lindorfflarsen.bsky.social
I don’t know but Don Knuth may have a partial answer:

“Email is a wonderful thing for people whose role in life is to be on top of things. But not for me; my role is to be on the bottom of things.”

cs.stanford.edu/~knuth/email...
Email (let's drop the hyphen) I have been a happy man ever since January 1, 1990, when I no longer had an email address. I'd used email since about 1975, and it seems to me that 15 years of email is plenty for one lifetime. Email is a wonderful thing for people whose role in life is to be on top of things. But not for me; my role is to be on the bottom of things. What I do takes long hours of studying and uninterruptible concentration. I try to learn certain areas of computer science exhaustively; then I try to digest that knowledge into a form that is accessible to people who don't have time for such study. On the other hand, I need to communicate with thousands of people all over the world as I write my books. I also want to be responsive to the people who read those books and have questions or comments. My goal is to do this communication efficiently, in batch mode --- like, one day every six months. So if you want to write to me about any topic, please use good ol' snail mail and send a letter to the following address: Prof. Donald E. Knuth Computer Science Department CoDa Building room W208 389 Jane Stanford Way Stanford University Stanford, CA 94305-5008 USA. I have a wonderful secretary who looks at the incoming postal mail and separates out anything that she knows I've been looking forward to seeing urgently. Everything else goes into a buffer storage area, which I empty periodically. My secretary also prints out all nonspam email messages addressed to taocp@cs.stanford.edu or knuth-bug@cs.stanford.edu, so that I can reply with written comments when I have a chance. If I run across such a message that was misaddressed --- I mean, if the message asks a question instead of reporting an error --- I try not to get angry. I used to just throw all such sheets in the wastebasket. But now I save them for scratch paper, so that I can print test material for The Art of Computer Programming on the blank sides. I DO NOT ANSWER UNSOLICITED EMAILS, nor do I respond t…
October 31, 2025 at 7:54 AM
lindorfflarsen.bsky.social
Accessible, uniform protein property prediction with a scikit-learn based toolset AIDE

Artificial Intelligence Driven protein Estimation (AIDE), enables instantiating, optimizing, and testing many zero-shot and supervised property prediction methods

doi.org/10.1093/bioi...
Figure 1 from the paper showing and overview of the AIDE package, which provides a scikit-learn model subclass that operates on a dataclass representing proteins as opposed to numpy matrices, and number of mixin protocols for protein models to adopt common behavior and compatibility checks for user data. These models are drop-in compatible with existing scikit-learn classes and pipelines, allowing for multistep processes to be defined within a single script in a reproducible way like one would for a traditional scikit-learn application. Table listing the methods currently available within AIDE
October 29, 2025 at 7:45 AM
lindorfflarsen.bsky.social
I attended a summer school on Spetses 🇬🇷 just as I was beginning my PhD. Looks like I will be going again in 2026 with "Lost in Integration vol. 2". Details to follow.
October 28, 2025 at 9:22 PM
lindorfflarsen.bsky.social
Just because this came up a few times recently

If you need to compare two or more ensembles (like RMSD for structures) we developed algorithms and code (ENCORE) to do just that

Similarity Measures for Protein Ensembles
journals.plos.org/plosone/arti...

ENCORE
journals.plos.org/ploscompbiol...
October 27, 2025 at 8:00 AM
lindorfflarsen.bsky.social
We (@sobuelow.bsky.social) developed AF-CALVADOS to integrate AlphaFold and CALVADOS to simulate flexible multidomain proteins at scale

See preprint for:
— Ensembles of >12000 full-length human proteins
— Analysis of IDRs in >1500 TFs

📜 doi.org/10.1101/2025...
💾 github.com/KULL-Centre/...
Figure showing the AF-CALVADOS restraining and simulation protocol based on AF2 structure, PAE and pLDDT
October 20, 2025 at 11:26 AM
lindorfflarsen.bsky.social
We thank reviewer 2 for their constructive suggestions and have updated our manuscript accordingly
October 15, 2025 at 10:42 PM
lindorfflarsen.bsky.social
Looking forward to visit @friedlab.bsky.social @jhu.edu on Tuesday Oct 14th to talk about

Prediction and Design of Intrinsically Disordered Proteins and Condensates 🍝💦

chemistry.jhu.edu/event/the-ep...
Title slide for the talk entitled Prediction and Design of Intrinsically Disordered Proteins and Condensates
October 12, 2025 at 2:51 PM
lindorfflarsen.bsky.social
One of the great advantages of biology is that you can write HELLO WORLD in amino acids since PROTEINS ARE AMAZING
October 9, 2025 at 11:51 AM
lindorfflarsen.bsky.social
Am at the University of Copenhagen Frontier Research in Denmark meeting celebrating the @erc.europa.eu. Rector David Lassen highlights how his first meeting with the ERC was a grant rejection, and how it's important to tell early career scientists about the CV of failure in addition to successes 1/n
October 8, 2025 at 8:06 AM
lindorfflarsen.bsky.social
The proteasome-substrate-shuttle protein UBQLN2 contains—like other quality control system proteins—a long region devoid of lysine (a lysine desert)

Martin Grønbæk-Thygesen (from @rhp-lab.bsky.social) et al show that introducing K here causes ubiquitylation and degradation

doi.org/10.1101/2025...
Fig. 1 – UBQLN2 is a conserved lysine-depleted protein. (A) Sequence comparison of UBQLN2 orthologs in the indicated species. Intrinsically disordered regions in human UBQLN2 based on MobiDB are shown as a blue bar. The domain organization based on the SMART database is marked. Lysine residues are marked as black squares. (B) ESM-2 predictions of all possible single amino acid substitutions of human UBQLN2 presented as a heat map. The wild-type residues are marked in blue. ESM-2 scores close to zero (light yellow colors) indicate that the amino acid substitution is compatible with the ESM-2 language model, whereas negative scores (dark orange colors) indicate that the variant is incompatible with the ESM-2 model. The domain organization (based on SMART) is aligned above the map. Note that substitutions to lysine or cysteine in general appear detrimental, in particular downstream of the UBL domain. (C) The AlphaFold2 predicted structured of human UBQLN2 (AF-Q9UHD9-F1) (left panel). The UBL domain is colored blue, and the UBA domain is colored orange and the STI1 regions green. Zoom in on the UBL domain (right panel) with the lysine residues highlighted as stick representations and colored based on the relative accessible surface area (rASA, dark red exposed; grey, buried).
October 7, 2025 at 9:20 PM
lindorfflarsen.bsky.social
A good day to remember John Gurdon’s school report from his biology master at Eton
October 7, 2025 at 8:45 PM
lindorfflarsen.bsky.social
Visiting Ephesus at #embointegmod25 to look for the missing AlphaFold 3 code
October 2, 2025 at 2:32 PM
lindorfflarsen.bsky.social
Integrative modelling of biomolecular dynamics

Time-dependent and -resolved experiments combined with computation provide a view on molecular dynamics beyond that available from static, ensemble-averaged experiments

Review w @dariagusew.bsky.social & Carl G Henning Hansen
doi.org/10.48550/arX...
Figure 1 from the review. Caption: Comparison of a schematic example showing static, time-dependent, and time-resolved experiments illustrated by a protein folding process. (a) A static experiment measuring the observable O$_{\text{exp}}$ is shown, which can be modelled as a distribution of simulated values, O$_{\text{calc}}$, representing a conformational ensemble of folded and unfolded states. (b) Shows a time-dependent experiment, where the equilibrium dynamics of reversible folding gives rise to measured transition times $\tau_1$ and $\tau_2$. These can be modelled as equilibrium dynamics, illustrated by a free energy (FE) surface along a chosen degree of freedom (D.O.F.) (c) A time-resolved experiment probes a non-equilibrium process, where the system begins at $t_{0}$ in the folded state. During the observation time $t$ the protein unfolds until $t_{\text{max}}$. At each time point, a distinct ensemble average, O$_{\text{exp}}$, can be observed, reflecting the proteins changing structure. This evolution can be modelled as distributions of O$_{\text{calc}}$ at each time point. These are shown together with a FE surface.
October 2, 2025 at 7:54 AM
lindorfflarsen.bsky.social
Thanks. I was looking to use this protocol github.com/agitter/ai-c...
from @anthonygitter.bsky.social to determine an author list, so needed something automated (and simple)
Screenshot from "A renewed call for open artificial intelligence in biomedicine" with the footnote: "All authors except A.G. were ordered arbitrarily using the mean confidence scores of protein structures predicted from their names"
October 1, 2025 at 1:49 PM
lindorfflarsen.bsky.social
Polishing my slides for #embointegmod25 in Izmir 🇹🇷. Perfect place to talk about integrative modelling of disordered and flexible proteins 🍝
Structure of an intrinsically disordered protein predicted by AlphaFold. The outline of the structure looks like a map of Turkey. The protein is Dentin Sialophosphoprotein (Uniprot: Q9NZW4) Map of Turkey
October 1, 2025 at 9:38 AM
lindorfflarsen.bsky.social
As it stands, the new rules (I think) prohibit HHMI authors from publishing in Science/AAAS and Cell press journals
September 25, 2025 at 6:14 PM
lindorfflarsen.bsky.social
Martin Grønbæk-Thygesen in @rhp-lab.bsky.social measured abundance and toxicity (both independently of function) of missense variants in aspartoacylase, and found that a subset of the low-abundance variants were toxic, induced a stress response which correlated with toxicity

doi.org/10.1038/s414...
September 19, 2025 at 8:57 PM
lindorfflarsen.bsky.social
The to-dodo list
September 17, 2025 at 4:34 PM
lindorfflarsen.bsky.social
New review on computational design of intrinsically disordered proteins 🖥️🍝 by @giuliotesei.bsky.social @fpesce.bsky.social & 👴

doi.org/10.48550/arX...
Figure 3 from the paper with the caption: "Role of machine learning in de novo design of IDRs. (A) Machine-learning models can be trained on diverse data sources, from molecular dynamics simulations to annotations of cellular localization and protein structures from the Protein Data Bank. (B) Often implemented as neural networks using sequence-encoded features as input, these models can initially be trained on a limited region of sequence space as surrogate models. Through active learning, additional simulations are performed during the design campaign to generate new data, and the surrogate model is retrained on the expanded dataset to progressively improve its accuracy. (C) Machine-learning models have been developed to predict biophysical observables, biological annotations, and protein structures. When combined, machine-learning models can be used to identify a set of sequences that strike a trade-off between multiple design objectives, defining a Pareto front."
September 17, 2025 at 4:17 PM
lindorfflarsen.bsky.social
This is in line with old data from Danish universities. Dark green shows expenses and light green shows revenues from commercialisation. As one example, the Danish Technical University spent 27M DKK and got 11M DKK back. Admittedly this is old data and may have changed.
From: dm.dk/media/ams/36...
Plot showing progress in expenses (dark green) and expenses (light green) for commercialisation from Danish Universities 2004–2015. Taken from https://dm.dk/media/ams/36084/ff-303.pdf
September 15, 2025 at 5:47 AM
lindorfflarsen.bsky.social
Interesting discussion between Reviewer 1 and the authors of "Redox-driven mineral and organic associations in Jezero Crater, Mars" and whether the paper will attract media attention as evidence of past Martian life
doi.org/10.1038/s415...
September 10, 2025 at 8:35 PM
lindorfflarsen.bsky.social
Back to teaching protein stability and including data from a wonderful paper showing why brewing beer at low-vs-high altitude affects the stability of the foam

Impact of Different Wort Boiling Temperatures on the Beer Foam Stabilizing Properties of Lipid Transfer Protein 1
doi.org/10.1021/jf03...
September 10, 2025 at 7:23 PM
lindorfflarsen.bsky.social
May the best team win 🇩🇰-🏴󠁧󠁢󠁳󠁣󠁴󠁿
September 5, 2025 at 6:26 PM