Lightnews — Scholar-powered news

Reposted by Umberto Lupo

Kresten Lindorff-Larsen @lindorfflarsen.bsky.social · 7d

Visiting Ephesus at #embointegmod25 to look for the missing AlphaFold 3 code

2 18

Reposted by Umberto Lupo

Lenka Zdeborová @zdeborova.bsky.social · 9d

Would you like to come to my lab for a postdoctoral position? Apply for the EPFL AI Centre fellowship: epfl.ch/research/fun.... FYI, I have funding (ERC & Simon's collaboration) also for 1-2 top candidates who will not secure the fellowship. Same for
@krzakalaf.bsky.social lab.

EPFL AI Center and Swiss AI Initiative Postdoctoral Fellowships

The 2nd call is now open with a deadline for submissions of 3 November (17.00 CET)!Applications are encouraged from researchers at the postdoctoral level with a keen interest in collaborative, interdi...

epfl.ch

1 2

Umberto Lupo @umbislupo.bsky.social · 23d

Very neat video, and equally neat housekeeping :)

1

Reposted by Umberto Lupo

Patrick Kidger @patrickkidger.bsky.social · 27d

💥 We are *also* organizing Machine Learning for Structural Biology @ EurIPS, Copenhagen!!

Topics include anything in the ML+bio intersection.

Submit your ML+bio short papers! Authors can even present in both locations if they have people in both locations 🚀

MLSB (in San Diego + Copenhagen) @workshopmlsb.bsky.social · 27d

You asked and we listened... @workshopmlsb.bsky.social is excited to be expanding to Copenhagen, DK at @euripsconf.bsky.social 🎉

Two workshops (San Diego & Copenhagen) will run concurrently to support broader attendance. You can indicate your location preference(s) in the submission portal💫

2 17

Umberto Lupo @umbislupo.bsky.social · Sep 4

Cool-looking work by @piompons.bsky.social @albecazzaniga.bsky.social @fra-cutu.bsky.social, going beyond existing works on sampling different conformational ensembles with AF2.

Valerio Piomponi @piompons.bsky.social · Sep 4

Our study is now published on JCIM🎉

We expanded and refined the preprint thanks to the insightful feedback from reviewers!

paper: pubs.acs.org/doi/10.1021/...
code: github.com/RitAreaScien...

Evolutionary Constraints Guide AlphaFold2 in Predicting Alternative Conformations and Inform Rational Mutation Design

Investigating structural variability is essential for understanding protein biological functions. Although AlphaFold2 accurately predicts static structures, it fails to capture the full spectrum of f...

pubs.acs.org

2

Reposted by Umberto Lupo

Nick Boyd @nboyd.bsky.social · Sep 2

Recently tested some de-novo minibinders against two targets (thanks Adaptyv!) designed using our open-source design library, `mosaic`; our best method got hit rates of 7/10 and 8/10 and affinities as low as single-digit nanomolar. Wrote up some thoughts here: blog.escalante.bio/minibinder-d...

2 15 38

Umberto Lupo @umbislupo.bsky.social · Sep 2

👍

Umberto Lupo @umbislupo.bsky.social · Sep 2

Is there any way, any way at all, that this can be followed remotely? (jk, but if you have a link...)

1

Reposted by Umberto Lupo

Nature @nature.com · Aug 29

Posts about research on Bluesky receive substantially more attention than similar posts on X, formerly called Twitter

go.nature.com/45Ftiw4

Research posts on Bluesky are more original — and get better engagement

Bluesky posts about science garner more likes and reposts than similar ones on X.

go.nature.com

2 110 300

Umberto Lupo @umbislupo.bsky.social · Aug 29

Two exciting openings with us! 🤖🧬🆎🧫💉
- AI Scientist 👉 lnkd.in/eDXHH4E8
- AI Scientist, Drug Creation 👉 lnkd.in/eEvGyaTR

You'll work on antibody sequence/structure design, antibody-antigen co-folding, antibody-antigen binding prediction, physics-based methodologies, and more!

DMs welcome!

2 3

Reposted by Umberto Lupo

Damiano Sgarbossa @damianosg.bsky.social · Aug 21

🎉 Excited to share that the last paper of my PhD is now published in PRX Life!

We introduce RAG-ESM, a retrieval-augmented framework that makes pretrained protein language models (like ESM2) homology-aware with minimal training cost.

📄 Paper: journals.aps.org/prxlife/abst...

Damiano Sgarbossa @damianosg.bsky.social · Apr 11

📢 Our new preprint is out on bioRxiv! We introduce RAG-ESM, a retrieval-augmented framework that improves pretrained protein language models like ESM2 by making them homology-aware with minimal additional training costs.
🔗 doi.org/10.1101/2025...
💻 github.com/Bitbol-Lab/r...

1/7

2 7

Umberto Lupo @umbislupo.bsky.social · Aug 21

@cyrilmalbranke.bsky.social is 🦋!

"ProteomeLM significantly outperforms DCA in recovering experimentally validated interactions. In H. sapiens, ProteomeLM achieves an AUC of 0.83, compared to 0.73 for DCA. Among the top 10 million scored pairs, it recovers 50% of known PPI, versus only 20% for DCA"

1

Umberto Lupo @umbislupo.bsky.social · Aug 21

Great to see this long-form version of your ICLR 2024 work finally out!

1

Reposted by Umberto Lupo

Paolo De Los Rios @paolodelosrios.bsky.social · Aug 21

A no-miss opportunity to join EPFL. We are a vibrant (had to use it 😉) community working in all branches of basic sciences and engineering, and interdisciplinary studies are strongly encouraged.
If you want more information, contact the committee heads, or feel free to DM me.

tinyurl.com/bd27ky7p

Assistant/associate Professor of AI-assisted Biophysics at the Ecole polytechnique fédérale de Lausanne (EPFL) and Group Leader at the Paul Scherrer Institute (PSI)

The School of Basic Sciences (Physics, Chemistry and Mathematics) at EPFL and the Paul Scherrer Institute (PSI) jointly seek to appoint a Tenure Track Assistant or tenured Associate Professor in AI-As...

tinyurl.com

18 21

Umberto Lupo @umbislupo.bsky.social · Aug 19

"When even the most benign activities are subject to capricious disruption and political interference, the luxury of disengagement is no longer a viable option."

Terence Tao @teorth.bsky.social · Aug 18

I wrote an op-ed on the world-class STEM research ecosystem in the United States, and how this ecosystem is now under attack on multiple fronts by the current administration: newsletter.ofthebrave.org/p/im-an-awar...

I’m an award-winning mathematician. Trump just cut my funding.

The “Mozart of Math” tried to stay out of politics. Then it came for his research.

newsletter.ofthebrave.org

Reposted by Umberto Lupo

Kieran Didi @kdidi.bsky.social · Aug 15

AtomWorks is out! Building upon @biotite_python, we built a toolkit for all things biomolecules and trained RF3 with it. All open-source, test it via `pip install atomworks`!

AtomWorks: github.com/RosettaCommo...
RF3: github.com/RosettaCommo...
Paper: tinyurl.com/y2w4z65b

1/6

1 6 21

Umberto Lupo @umbislupo.bsky.social · Aug 10

Thanks, that makes a lot of sense!

1

Umberto Lupo @umbislupo.bsky.social · Aug 9

(I.e., each antitoxin mutant is evaluated by making a single-row concatenated MSA with the fixed toxin?)

1

Umberto Lupo @umbislupo.bsky.social · Aug 9

Small question about this point: here, what input MSA(s) did you use? I'm guessing it was concatenated MSAs with just one row?

1

Reposted by Umberto Lupo

Noah Giansiracusa @profnoahgian.bsky.social · Jun 4

From social media apps to bank loans and hiring/firing decisions to search engines and shopping sites, are you tired of living a life ruled by opaque algorithms?

Pre-order now to Take Back Control on August 5.
bit.ly/4idpeGr

1 3

Umberto Lupo @umbislupo.bsky.social · Aug 5

One could (modulo depth considerations) use the total/mean MSA perplexity of the concatenated MSAs under MSA Pairformer. The pairing with lowest MSA perplexity should be favoured?

1 2

Umberto Lupo @umbislupo.bsky.social · Aug 5

*for all proteins (regardless of taxon or encoding location)

1

Umberto Lupo @umbislupo.bsky.social · Aug 5

Ah-ha, thanks for the clarification! I am mostly wondering how much of this was unexpected and, relatedly, whether you have decided to turn this pairing mode on for all taxa (and why).

1

Reposted by Umberto Lupo

Yo Akiyama @yoakiyama.bsky.social · Aug 5

Excited to share work with
Zhidian Zhang, @milot.bsky.social, @martinsteinegger.bsky.social, and @sokrypton.org
biorxiv.org/content/10.1...
TLDR: We introduce MSA Pairformer, a 111M parameter protein language model that challenges the scaling paradigm in self-supervised protein language modeling🧵

Scaling down protein language modeling with MSA Pairformer

Recent efforts in protein language modeling have focused on scaling single-sequence models and their training data, requiring vast compute resources that limit accessibility. Although models that use ...

biorxiv.org

1 43 95

Umberto Lupo @umbislupo.bsky.social · Aug 5

How much are you surprised by the case of 2Y69? It being a eukaryote, while I believe 3A0R is less surprising.

1