Sign in

Yousuf A. Khan

@yousufakhan.bsky.social
1.5K followers 259 following 76 posts

Group Leader @Stanford. RNA focused ML/AI, cryoEM/ET, and biophysics. Formerly:DeepMind AlphaFold,EvoscaleAI, Churchill Scholar@Cambridge_Uni & seen on Netflix

PostsRepliesMedia
Yousuf A. Khan @yousufakhan.bsky.social · 21/09/2026
Great story Daniel on NS3 and exciting to see the different states these RNA AAA+ complexes can adopt
120
Reposted by Yousuf A. Khan
Martin Steinegger 🇺🇦 @martinsteinegger.bsky.social · 01/08/2026
Riboseek is a fast RNA/DNA search. More sensitive than nhmmer at 250x speed. Structure-aware realignment produces MSAs approaching rMSA quality. Plus 1.7M precomputed RNA MSAs, and an API to search your own 📄 www.biorxiv.org/content/10.6... 💾 github.com/steineggerla... 🌐 search.foldseek.com/riboseek
218075
Yousuf A. Khan @yousufakhan.bsky.social · 16/06/2026
Realized I forgot to use #CryoEM tag to disseminate our cool structure of the 2C:RNA holoenzyme complex !
061
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
Thank you Daniel, excited to see the NS3 work in its published form!
010
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
Biorxiv: www.biorxiv.org/content/10.6...
biorxiv.org
The Structure of the Picornaviral 2C:RNA holoenzyme: Molecular Basis of RNA binding and specificity by a AAA+ protein
Picornaviruses are one of the leading agents of animal and human infectious disease with at least 8 billion infections a year and cause a range of symptoms including respiratory failure and acute flac...
031
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
This work could not have been possible without the amazing support of our collaborators in Dr. Trevor Sweeney’s laboratory and of course the amazing people (Richard, Louie and Maeva) that accelerated this work so much (9/n)
110
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
Lastly, we used structural phylogenetics to search through metagenomic databases and found that there are many 2C-like proteins in many different viral families outside of picornaviruses, suggesting that our work here could have even broader ramifications than we thought (8/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
We found that this AAA+ 2C holoenzyme more resembles a protein translocase rather than a helicase, which was unexpected but then again, this is the first ever RNA AAA+ complex ever seen. (7/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
Uniquely, it uses carbonyl atoms from the backbone of one of its pores to confer specific RNA selectivity hydrogen bonding at the 2′-OH. This reminded us so much of how ion channels confer selectivity to substrate by also using backbone carbonyls in a pore (6/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
In our work, we characterize the unique and specific binding modality of this AAA+ protein bound to RNA. No AAA+ protein has ever been seen to bind to ssRNA, so this work also provides the biophysical principles of how this protein specifically binds RNA (5/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
Much work has been devoted to developing inhibitors and therapeutics targeting 2C, specifically in enteroviruses that have caused several outbreaks all around the world recently, but no structure of 2C bound to its substrate RNA has ever been solved until now (4/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
2C is one of the most highly conserved proteins in all of these different viruses (e.g. enteroviruses, rhinoviruses, hepatitis A virus, poliovirus etc) and has been known to bind to RNA as a critical complex in viral replication (3/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
Picornaviruses cause likely upwards of 8 billion infections per year and pose a severe disease and death toll on mankind. Beyond that, picornaviruses can also infect cattle and overall pose a severe economic burden to the glove as well (2/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 11/06/2026
Today my laboratory is thrilled to announce a major breakthrough in the fight against the common cold and other diseases like hand,foot, and mouth disease acute flaccid myelitis caused by picornaviruses. We present the first structure of picornaviral 2C, bound to ssRNA (1/n)
2158
Yousuf A. Khan @yousufakhan.bsky.social · 01/06/2026
github.com/YousufAKhan/...
000
Yousuf A. Khan @yousufakhan.bsky.social · 01/06/2026
Great to see such a spike in interest for the RNA design capabilities of yakRNA Design after my talk at @rnasociety.bsky.social . Thank you to everyone for the lovely feedback to me and the rest of the team! I'm so excited to see what people create with it!
100
Reposted by Yousuf A. Khan
Dimitry Tegunov @dtegunov.bsky.social · 03/05/2026
We're super excited to share MissAlignment: a new ML-based approach to reference-free tilt series alignment, spearheaded by @martenchaillet.bsky.social. We think it's going to make your cryo-ET life a lot better. Preprint: www.biorxiv.org/content/10.6... Code: github.com/warpem/miss-... 🧶 Thread:
6313118
Yousuf A. Khan @yousufakhan.bsky.social · 24/04/2026
Check out our preprint: biorxiv.org/content/10.6... And our GitHub where we have made the weights and inference code available along with an easy to use Google Colab that requires no installation: github.com/YousufAKhan/... (8/n)
biorxiv.org
000
Yousuf A. Khan @yousufakhan.bsky.social · 24/04/2026
One of our efficient synthetic RNAs had no homology to anything we could BLAST, or use a covariance model on (as compared to our other designs which just had many mutations away from a related sequence) underscoring how creative yakRNA design truly is (7/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 24/04/2026
This led to a model that can take any combination of these inputs to generate synthetic and functional sequences. Most impressively, we had yakRNA Design compose 84 synthetic RNA pseudoknots and (WITHOUT ANY SCORING/SELECTIOn), 17 of them were highly functional (6/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 24/04/2026
This semantically labeled dataset (semRfam) has, for each entry a sequence, a corresponding biological function, consensus, and a secondary structure that we challenged our model to co-learn between in a unified vocabulary (5/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 24/04/2026
To sidestep this completely, we went to one of the most well curated and annotated databases, Rfam. But instead of just pulling the raw sequences, we took the biological function descriptions, family information, secondary structure info all together (4/n)
110
Yousuf A. Khan @yousufakhan.bsky.social · 24/04/2026
This has led to groups trying to exploit sequencing databases but just training on large, unnannotated corupses of sequence can only be so effective and trying to co-train models with lots of sequences and few structures has not been successful (3/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 24/04/2026
The issue with RNA structure prediction and subsequently RNA design is that we just simply do not have enough training data. RNA only structures represent only >1% of the PDB, which leads to brittle models that are unable to generalize (2/n)
100
Yousuf A. Khan @yousufakhan.bsky.social · 24/04/2026
Introducing yakRNA Design🚀, an RNA composer that makes synthetic, programmed RNAs at the touch of your fingers! The crazy part? Unlike other macromolecular design models, we didn't use ANY 3D RNA structures in our model🤯 (1/n)
100
Reposted by Yousuf A. Khan
Mohammed AlQuraishi @moalquraishi.bsky.social · 13/03/2026
New OpenFold3 preview out! (OF3p2) It closes the gap to AlphaFold3 for most modalities. Most critically, we're releasing everything, including training sets & configs, making OF3p2 the only current AF3-based model that is functionally trainable & reproducible from scratch🧵1/9
124590
Yousuf A. Khan @yousufakhan.bsky.social · 13/03/2026
Check out our lab's latest "Journal Club" in Nature Review Genetics today where we talk about how molecular mechanisms can dynamically recode our genomes: www.nature.com/articles/s41...
nature.com
Recoding genomes with programmed ribosomal frameshifting - Nature Reviews Genetics
In this Journal Club, Yousuf A. Khan recounts a landmark study published in 1985 by Jacks and Varmus that documented programmed ribosomal frameshifting in retroviruses, which later inspired Yousu...
000
Yousuf A. Khan @yousufakhan.bsky.social · 03/12/2025
It's all about RNA baby! Ribose found on Bennu www.nature.com/articles/s41...
nature.com
Bio-essential sugars in samples from asteroid Bennu - Nature Geoscience
Samples returned from asteroid Bennu contain bio-essential sugars such as ribose and glucose that may have formed in the parent asteroid from brines containing formaldehyde, according to a geochemical...
020
Yousuf A. Khan @yousufakhan.bsky.social · 24/10/2025
This paper was a tour de force and would not be possible without he help of so many collaborators like @juanbonifacino.bsky.social and many more. Thank you!!!
011
Yousuf A. Khan @yousufakhan.bsky.social · 24/10/2025
Strikingly, the presence of this frameshifted proteoform was necessary for myocardial contraction, as we showed in PLEKHM2-KO cells that required both the WT and frameshifted proteoforms to recover fully, leading to our model
110
Yousuf A. Khan @yousufakhan.bsky.social · 24/10/2025
Furthermore, we were able to show that this frameshifted version of PLEKHM2 generated a hyper-active version of this protein, which did not require activation of its canonical activator
110
Yousuf A. Khan @yousufakhan.bsky.social · 24/10/2025
Not only did the frameshift exist, it was highly conserved and encoded for an additional alpha-helix domain
110
Yousuf A. Khan @yousufakhan.bsky.social · 24/10/2025
Today our journey ends! We were able to find a highly conserved cellular +1 PRF signal in the gene PLEKHM2/SKIP
110
Yousuf A. Khan @yousufakhan.bsky.social · 24/10/2025
A few years ago, we published in Nature (nature.com/articles/s41...) detailing this artefact and noting that no true PRF signal existed in humans gave access to two overlapping open reading frames that both encoded something useful like in viruses
110
Yousuf A. Khan @yousufakhan.bsky.social · 24/10/2025
However the authors (nature.com/articles/nat...) and others in the field had unfortunately been using a reporter with a malicious splicing artefact, accidentally presenting PRF activity.
110
Yousuf A. Khan @yousufakhan.bsky.social · 24/10/2025
PRF is employed widely across viral clades from HIV-1, Sars-COV-2, RSV, Influenza, etc. However, there was always a chase to find the first real case of PRF occurring in a human, cellular gene. And in 2014, researchers believed that had found the first case in CCR5
110
Yousuf A. Khan @yousufakhan.bsky.social · 24/10/2025
Programmed Ribosomal Frameshifting (PRF) is a phenomena that has been studied extensively in viruses since the 80s. In short, a single RNA has the ability to code for multiple proteins by 'recoding' the genetic code i.e. having the ribosome change its reading frame mid-elongation
120
Yousuf A. Khan @yousufakhan.bsky.social · 24/10/2025
Excited to announce the YAK lab's first paper and the discovery of the FIRST human cellular PRF signal to give access to two overlapping open reading frames (science.org/doi/10.1126/...)! Before we dive in, the story actually begins in a Nature from 11 years ago
194
Yousuf A. Khan @yousufakhan.bsky.social · 13/10/2025
Excited to announce that I have been selected to receive a High-Risk, High-Reward grant from the NIH to help support the research our lab does at the intersection of AI and molecular biology to study RNA. Read more about here: news.stanford.edu/stories/2025...
news.stanford.edu
Stanford researchers receive NIH High-Risk, High-Reward grants
This year’s awardees are researching cancer treatments, age-related disease, and much more.
030
Yousuf A. Khan @yousufakhan.bsky.social · 27/08/2025
Excited to collaborate and have people join us! Please checkout the lab website above if you are interested!
000
Yousuf A. Khan @yousufakhan.bsky.social · 27/08/2025
I'm extremely thrilled to announce that I will be starting as a group leader at Stanford this upcoming fall! My lab will be studying how RNAs exert themselves in the cell to induce a multitude of biological phenomena using ML, cryoEM/ET, and biochemistry! www.yousufakhan.com
yousufakhan.com
Yousuf A. Khan (YAK) Lab - RNA Research at Stanford University
Yousuf A. Khan (YAK) Lab - Studying RNA folding, interactions, and cellular effects using Molecular and Cellular Physiology, Structural Biology, and Machine Learning at Stanford University
3174
Reposted by Yousuf A. Khan
Ahmad Jomaa @jomaalab.bsky.social · 01/08/2025
Check out our new preprint on the discovery of a molecular switch in NAC that mediates nascent chain sorting on the ribosome and prevents mitochondrial protein mistargeting by SRP. A great collaboration with the Shan Lab @Caltech and the Qi Lab @UVA: www.biorxiv.org/content/10.1...
48328
Yousuf A. Khan @yousufakhan.bsky.social · 29/07/2025
Don’t leave us hanging Hiten!
100
Reposted by Yousuf A. Khan
Alex Rubinsteyn @alexr.bsky.social · 23/07/2025
Going deep on 2A peptides: www.cell.com/cell-reports... (Trying to pick between GSG:T2A, GSG:P2A, and furin site:GSG:T2A)
cell.com
Systematic identification and characterization of eukaryotic and viral 2A peptide-bond-skipping sequences
Rao et al. identified thousands of previously unknown 2A peptides across both viruses and eukaryotes using an HMMER analysis. The authors further identified a unique class of 2A peptides, class B, who...
12511
Reposted by Yousuf A. Khan
Sergey Ovchinnikov @sokrypton.org · 22/07/2025
PSA = Google Colab Pro free for one year for academic use (US only): blog.google/outreach-ini...
blog.google
New Google Colab features for higher education
Google Colab offers free Colab Pro for students, interactive slideshows and AI controls in notebooks.
0227
Reposted by Yousuf A. Khan
Nikolai Slavov @slavov-n.bsky.social · 09/07/2025
Key numbers in cell biology Having a sense of scale helps to think more rigorously and realistically about biological systems.
010329
Reposted by Yousuf A. Khan
Nature Structural & Molecular Biology @natsmb.nature.com · 03/07/2025
ICYMI: New online: SNARE disassembly requires Sec18/NSF side loading
go.nature.com
SNARE disassembly requires Sec18/NSF side loading
Nature Structural & Molecular Biology, Published online: 02 July 2025; doi:10.1038/s41594-025-01590-wKhan et al. show that Sec18 uses both D1 and D2 AAA+ rings cooperatively, opening them laterally for SNARE substrate loading and for subsequent release of SNAREs.
042
Reposted by Yousuf A. Khan
Hiten Madhani @hitenmadhani.bsky.social · 02/07/2025
Cool paper on bacterial amyloids as protection from predatory bacteria from the Whitely lab at CU Boulder. There are predatory fungi, so one wonders if a similar system exists in eukaryotes. www.nature.com/articles/s41...
nature.com
Functional amyloid proteins confer defence against predatory bacteria - Nature
Escherichia coli uses curli fibres, oligomers of the functional amyloid CsgA, as a barrier to protect against the predatory bacteria Bdellovibrio bacteriovorus and Myxococcus xanthus in a mechanism th...
0408
Yousuf A. Khan @yousufakhan.bsky.social · 02/07/2025
Thus, any kind of SNARE can be processed universally by this single molecule machine. Check out the paper for more in-depth structural work! (5/5)
020
Yousuf A. Khan @yousufakhan.bsky.social · 02/07/2025
Combining in-vivo mass-spec, single molecule FRET, and time resolved cryoEM, we were able to determine that this NSF/Sec18 ring actually splits itself open from the side to accomodate SNAREs, thus completely bypassing any substrate topology issues! (4/5)
172