@stuartmac44

Computational biologist | Protein structures & human variation | R & Python | Genetic effects on protein function | See: https://xitter.cf/t.co/wzvNj7zDKI

 Dundee, Scotland
dundee.ac.uk/people/stuart-macgowan
Joined September 2012
  • Tweets242
  • Following462
  • Followers413
17Media
Stuart MacGowan retweeted
AI Frontiers
@ai_frontiers_15 May 2025

The Misguided Quest for Mechanistic AI Interpretability

Article · @ai_frontiers_
8
28
11
204
79,837
The github repo is still in development but available at github.com/jtrinquier/SoftAl….

GitHub - jtrinquier/SoftAlign

Contribute to jtrinquier/SoftAlign development by creating an account on GitHub.

github.com
1
2
367
Stuart MacGowan retweeted
Eric Topol
@EricTopol15 May 2025
Today "a milestone in the evolution of personalized therapies for rare & ultra-rare inborn errors of metabolism" —the 1st human to undergo custom genome editing —outgrowth of decades of NIH funded research nejm.org/doi/full/10.1056/NE… nejm.org/doi/full/10.1056/NE… @NEJM nytimes.com/2025/05/15/healt…
32
597
50
2,172
170,363
ProteinGuide: Predictor guidance for generation with any autoregressive or discrete diffusion protein model. @junhaobearxiong @jlistgarten
2
8
1
86
6,981
Stuart MacGowan retweeted
David R. Liu
@davidrliu15 May 2025
In a medical milestone, a customized base editor was developed, characterized in human and mouse cells, tested in mice, studied for safety in non-human primates, cleared by @US_FDA for clinical trial use, manufactured as a complex with an LNP, and dosed into a baby with a severe, rapidly progressing genetic disease... all in an astounding 7 months. Best of all, the infant patient shows apparent benefit. Congratulations to @kiranmusunuru, Rebecca Ahrens-Nicklas, and other team members for this heroic and inspiring effort, which has implications for the hundreds of millions of patients that suffer from thousands of genetic diseases. drive.google.com/file/d/1Jfk…
44
430
59
1,638
248,415
Stuart MacGowan retweeted
Sam Blau@SamMBlau14 May 2025
The Open Molecules 2025 dataset is out! With >100M gold-standard ωB97M-V/def2-TZVPD calcs of biomolecules, electrolytes, metal complexes, and small molecules, OMol is by far the largest, most diverse, and highest quality molecular DFT dataset for training MLIPs ever made 1/N
4
102
8
424
44,511
Stuart MacGowan retweeted
Angelica Parente
@draparente13 May 2025
If I wanted to design a mini binder protein that binds to a known small molecule, what’s the SotA model to use right now?
11
10
1
99
22,543
Stuart MacGowan retweeted
Ben Lehner@BenLehner4 May 2025
New publication. We quantify the aggregation of >100,000 random protein sequences to train CANYA, a convolution-attention hybrid neural network to predict aggregation from sequence. With @Bennibolo @CRGenomica @sangerinstitute @IBECBarcelona science.org/doi/10.1126/scia…

Massive experimental quantification allows interpretable deep learning of protein aggregation

xAI trained on >100,000 random peptides predicts and elucidates primary sequence determinants of aggregation.

science.org
2
36
1
150
13,546
Stuart MacGowan retweeted
Ben Lehner@BenLehner6 May 2025
Come and join us! We’re hiring a new Group Leader in Generative Biology at the @sangerinstitute Building AI models or the data to train them? Core funding of >$130M a year for a faculty of ~30. pls RT! nature.com/naturecareers/job… acrobat.adobe.com/id/urn:aai…
1
25
58
7,209
Stuart MacGowan retweeted
MindBranches
@MindBranches3 May 2025
Book Summary: “Unleash the Power of Storytelling” tap, hold, load 4k
6
106
1
918
93,224
MolSnapper has now been published in @JCIM_JCTC! MolSnapper integrates expert knowledge into diffusion models for structure-based drug design using a conditioning approach. Congratulations @YaeliZiv, @fergus_imrie, Brian Marsden, and Charlotte Deane. pubs.acs.org/doi/10.1021/acs…

MolSnapper: Conditioning Diffusion for Structure-Based Drug Design

Abstract. Generative models have emerged as potentially powerful methods for molecular design, yet challenges persist in generating molecules that effectiv

pubs.acs.org
8
38
1,483
We're hiring a Research Software Engineer to join OPIG and @OxfordStats! This is a permanent role to support the group's world-leading open source tools Grade 8 salary band £48k-£57k pa. Apply by 4th June 2025, noon UK time. Further details here: my.corehr.com/pls/uoxrecruit…
1
2
3
297
Happening today at **4:30PM** ET!
Machine Learning for Protein Engineering Seminar@ml4proteins22 Apr 2025
Next Tues (4/29) at **4:30PM** ET, we will have @ginaelnesr @HWaymentSteele present "Learning millisecond protein dynamics from what is missing in NMR spectra" Paper: biorxiv.org/content/10.1101/… Sign up on our website for zoom links!

Learning millisecond protein dynamics from what is missing in NMR spectra

Many proteins’ biological functions rely on interconversions between multiple conformations occurring at micro-to millisecond (µs-ms) timescales. A lack of standardized, large-scale experimental data...

biorxiv.org
3
9
3,400
Stuart MacGowan retweeted
Tom Dörr
@tom_doerr25 Apr 2025
Python library that adds interactive dashboards to pandas dataframes for exploring data
4
96
3
847
59,127
Complete view of de novo mutation rate across the genome estimated from a four-generation pedigree using five complementary short and long read sequencing technologies. The authors estimate 98–206 DNMs per transmission: - 74.5 SNVs - 7.4 tandem repeats - 65.3 structural variants - 4.4 centromeric DNMs Among males, 12.4 de novo Y-chromosomal mutations per generation. 16% of the de novo SNVs are post zygotic in origin. Impressive data from Eichler's lab! Porubsky et al. Nature nature.com/articles/s41586-0…
4
47
2
132
14,745
Stuart MacGowan retweeted
Sumanth
@Sumanth_07723 Apr 2025
Microsoft launched the best course on Generative AI! The free 21 lesson course is available on Github and will teach you everything you need to know to start building Generative AI applications.
53
458
19
2,284
207,438
A curated list of free courses with certifications.
10
356
9
2,964
192,932
Stuart MacGowan retweeted
Docker
@Docker22 Apr 2025
Coming soon! We’re introducing the Docker MCP Catalog and ToolKit to streamline how developers discover, install, authenticate, and connect #MCP servers to their favorite clients. youtube.com/watch?v=98M_6njO…

Announcing Docker MCP Catalog and Docker MCP Toolkit

Coming soon! We’re introducing the Docker MCP Catalog and ToolKit t...

youtube.com
3
33
6
124
27,751
My favourite is “Plateaus are a sign to evolve your approach, not a signal to quit.”
MindBranches
@MindBranches19 Apr 2025
Book Summary: “The Science of Rapid Skill Acquisition” tap, hold, load 4k
49
We have released the successor to ANARCI - ANARCII - a suite of Seq2Seq language models trained to number antibody (or TCR) sequences! Read the paper: biorxiv.org/content/10.1101/… Play with the webtool: opig.stats.ox.ac.uk/webapps/… Documentation and codebase: github.com/oxpig/ANARCII

ANARCII: A Generalised Language Model for Antigen Receptor Numbering

Antigen receptor numbering allows the rapid delineation of the antigen-binding regions of antibody and T cell receptor (TCR) sequences, from sequence alone. It also allows the comparison of the vast...

biorxiv.org
1
6
27
1,383
Load more