Predict protein structure and function using sequence, evolution & more! Operated by lab members.
Munich & New York
Joined November 2010
- Tweets499
- Following73
- Followers733
- Likes40
RostLab retweeted
Very happy to introduce VespaG, for ultra-fast and accurate prediction of protein mutational effects. github.com/JSchlensok/VespaG. Big thanks to @LiniMuc, Julius Schlensok, and @MarinaAbakarova for the great work and effort. Exciting collab with @rostlab 1/5
VespaG: Expert-guided protein Language Models enable accurate and blazingly fast fitness prediction biorxiv.org/cgi/content/shor… #biorxiv_bioinfo
RostLab retweeted
Our new bilingual protein language model (pLM), ProstT5, translates between protein sequence and structure. Besides producing more structure-aware embeddings that are better at remote homology detection than sequence-pLMs, its translation capability enables inverse folding.
RostLab retweeted
We (@lindner_seb, @HeinzingerM) introduce REXzyme, a new model for the generation of enzymes that catalyze user-defined chemical reactions. We'd like that the community can benefit as soon as possible from it, so we made it available in HuggingFace: huggingface.co/AI4PD/REXzyme.
RostLab retweeted
Is ColabFold's MMseqs2-based MSA generation protocol suitable for mutational outcome prediction? Yes! Joint work by @MarinaAbakarova and Céline Marquet. 1/2
Alignment-based protein mutational landscape prediction: doing more with less biorxiv.org/cgi/content/shor… #bioRxiv
RostLab retweeted
Since the # of "[α-ω](fold){0,1}[A-z]*" tools are getting out of hand, I tried to curate a lil list of the ones I'm aware of: github.com/sacdallago/foldin…
Peeps w/ overview @sokrypton @thesteinegger @KevinKaichuang feel free to contribute if I missed stuff?
RostLab retweeted
LambdaPP: Fast and accessible protein-specific phenotype predictions biorxiv.org/cgi/content/shor… #bioRxiv
RostLab retweeted
PredictProtein computes a wide range of predictions of function and structure from sequence in max 8 seconds. EIGHT!
With the new release of AFDB, you can now input a UniProt identifier and get the AFDB structure + predicted features almost instantly:
embed.predictprotein.org/#/A…
RostLab retweeted
A proud moment when @HannesStaerk and Céline propose an amazing project for @TU_Muenchen students: mdsi.tum.de/di-lab/projekte/…
These two have become so much better than I at whatever they set out to do that it's only good they train the next wave of scientists.
RostLab retweeted
A collection of articles on machine learning for protein science and engineering: who would you like to hear from and on what topics?
RostLab retweeted
Contrastive learning using CATH hierarchies on protein language model embeddings -> fast, MSA-free search for distant homologs.
Michael Heinzinger, Maria Littmann, Nicola Bordin, Christine Orengo, @rostlab
biorxiv.org/content/10.1101/…
RostLab retweeted
This was sooooooooo much fun to write!!!
Remember when I mentioned my PI signing what I thought was Gwen Stefani, but in reality was fiddler on the roof: youtu.be/RBHZFYpQ6nc
How are these things connected? A hint is in the title, but for the full pic you gotta join @rostlab
This Post is from an account that no longer exists. Learn more
RostLab retweeted
The v2 of our localization work is pre-printed, finally with much needed comparisons to other protein sequence representation work & updated "lessons learned" along the way: biorxiv.org/content/10.1101/…
RostLab retweeted
The SARS-CoV-2 structural coverage map, led by Sean O’Donoghue @GarvanInstitute, co-authored by
@rostlab & others is out now @MolSystBiol doi.org/10.15252/msb.2020100…
Good tooling (aquaria.ws + predictprotein.org + ...) enables big pic analyses of bio mechanisms!
RostLab retweeted
>2,000 structural models spanning 69% of the #SARSCoV2 proteome, in a visual layout that summarises current knowledge on viral protein structures & provides insights into viral replication, mimicry, hijacking --> bit.ly/3hwALUG
S O’Donoghue @GarvanInstitute @rostlab
RostLab retweeted
Protein embeddings take over the world: now predict residues to bind metal ions, nucleic acids, or small molecules! Lead by one of the most consequential scientists I know (Maria Littmann) @rostlab
Protein embeddings and deep learning predict binding residues for various ligand classes biorxiv.org/cgi/content/shor… #bioRxiv
RostLab retweeted
PredictProtein, my 1-month younger scientific brother, is now out in NAR (doi.org/10.1093/nar/gkab354). This was a tremendous amount of fun and a great way to bridge the traditional to the modern (teapawt.net/sacdallago/status/1392…), with so many more exciting things to come in the near future!
bio_embeddings (bioembeddings.com) is now out as a protocol (doi.org/10.1002/cpz1.113)! A suite of "things" to make bio LMs (and whatever's built on top) more accessible to everyone (biologist, machine learner, MLOps-er, software engineer, cluster or laptop user).
RostLab retweeted
bio_embeddings (bioembeddings.com) is now out as a protocol (doi.org/10.1002/cpz1.113)! A suite of "things" to make bio LMs (and whatever's built on top) more accessible to everyone (biologist, machine learner, MLOps-er, software engineer, cluster or laptop user).
RostLab retweeted
Today, I'm happy to announce that we were able to reach the top for protein sub-cellular location prediction using single sequence inputs from protein language models. A first (being first!) in function prediction from embeddings! 1/
Light Attention Predicts Protein Location from the Language of Life biorxiv.org/cgi/content/shor… #bioRxiv
RostLab retweeted
Are you working on representation learning in biology? You might be interested in attending the special session on representation learning in biology at ISMB/ECCB 2021 representation.learning.bio
Soon: submit your own research for lightning talks and poster presentations!
RostLab retweeted
PredictProtein - Predicting Protein Structure and Function for 29 Years biorxiv.org/cgi/content/shor… #bioRxiv