Every grant I applied for had an unprecedented number of applications this year, while simultaneously there is less overall funding. Very discouraging when putting in the effort no longer matters and blasting the system with AI written grants becomes the winning strategy.
Papers are like buses... You wait for ages, then two come along at once.
Huge congrats to @bornanovak.bsky.social and @jefflotthammer.bsky.social for pushing and driving every aspect of this work, preprinted ~1 year ago to the day (Friday before BPS), now published!
www.nature.com/articles/s41...
Pleased to share the final version of this behemoth of a paper, now finally published. I guess I can retire now?
www.nature.com/articles/s41...
More functional data, many thousands of words removed, and a few other updates from last year's preprint.
Computer science needs more biologists. They are used to:
- study systems that are mindbogglingly complex and complicated,
- use reductionism to study and describe such systems, using famously clumsy and fiddly tools to collect noisy and incomplete data,
(1/2)
AlphaGenome is out in @nature.com today along with model weights! 🧬
📄 Paper: www.nature.com/articles/s41...
💻 Weights: github.com/google-deepm...
Getting here wasn’t a straight path. We discussed the story behind the model, paper & API in the following roundtable: youtu.be/V8lhUqKqzUc
Agreed! Great we have this model & kudos to the developers (especially for user-friendly implementation😀), but let's not abuse it as an excuse to stop trying to understand the actual mechanisms! Yes, gene regulation is complicated, but still rooted in the biophysics of molecular interactions...
Surprised (but also not that surprised) that the AlphaGenome paper didn't officially cite any of the primary data used for training their model (see Fig. 1, thousands of datasets made with tremendous time and effort over >15yrs). What's up with that @nature.com ?
www.nature.com/articles/s41...
My next read!
Great collaboration! 📠+🧬 Experiments and ML go hand-in-hand. Hopefully more to come in the future!
"Our findings argue against a strict reg. role for TAD boundaries in controlling gene expression [...] TADs appear to serve primarily to organize chrom. topology at a global scale"- taking it as support from the '3Ders' for my love of TFs as prime regulatory suspects😉
www.biorxiv.org/content/10.6...
Overview of Integrated Robotic Imaging and Sequencing, IRIS
We just released IRIS (7+yrs project), a tech we believe will transform cell biology by pairing high-resolution cell images with matched #scRNAseq, letting us interpret cellular form by its molecular ground truth.
Huge tx to @JohannesBues, @JoernPezoldt, @CamilleLambert et al.
shorturl.at/zgY8Z
Very much into this! "Future work will clarify the biochemical basis for how promoter-proximal elements, like SRR2, synergize with distal enhancers to support cohesin-independent long-range regulation."
Now that I'm settled in at @umasschan.bsky.social, I'm hiring at all levels: grad students, post-docs, and software engineers/bioinformaticians!
The goal of my lab is to understand the regulatory role of every nucleotide in our genomes and how this changes across every cell in our bodies.
While over here discussions are about slashing Science funding, the Swiss are proposing to put it center stage on their currency. And I thought I was returning to the "Mecca of Science", not leaving it... 😅 Also, is this even real money? www.neuebanknotenserie.ch/detailview.h...
Beautifully controlled system to delineate primary from context-specific TF binding, revealing how chromatin state shapes TF binding specificity during differentiation and vice versa!
Excited to see that our study on context-specific transcription factors was named as one of 8 remarkable outputs of 2024 by the Swiss Institute of Bioinformatics.
The Range Extender
For enhancers to function
At long distances
#ChromatinHaiku
Congrats!
same here. Have an 18-month-old and the only time I get anything done is at night when he sleeps 🫠 You are not alone, hang in there!
Agree, there are definitely cases where syntax is important, the question is more how to model it well when it isn't. I would think the latter is important for accuracy, which is where models tend to start underperforming (e.g. predicting activity for personalised genomes). I like your suggestion!
might be time to reconsider how we train machine learning models. Almost every model I come across uses convolutional filters that attempt to learn syntax, when perhaps none is present...
Really cool work using a gene synteny-based strategy to identify enhancer conservation across species w/o seq similarity. Last figure is my favorite: while TFBS content is conserved, TFBS syntax is not! Aligns with our findings on TF cooperativity, where a specific syntax was not required.
As a scientist and mom of a little boy this story really touched me. The rewards of science funding go way beyond $ return on investment (& NIH's ROI is unparalleled). Hats off to the parents making this incredibly tough decision, paving the way for future treatments. www.nytimes.com/2025/05/15/h...
How is it end of April already? Excited to share that the Kribelbauer Lab is up and running @USC/MCB. Grateful for my first two PhD students @m-finegan.bsky.social @christinagirgis.bsky.social. We are dev scalable, genome-int. tools to study TF & enhancer regulation. Website now live bit.ly/4jo8Tjg
MPRAbase (mprabase.ucsf.edu) , a customized database for massively parallel reporter assays (MPRAs) to easily find and download MPRA data. Amazing work by Jingjing Zhao, Fotis Baltoumas, Georgios Pavlopoulos, @vagar.bsky.social, ilias Georgakopoulos-Soares & others.
genome.cshlp.org/content/earl...
100% agree w Lowe on this: "But how is AGI going to suddenly reveal what is now hidden? The sum total of all the medical information in the world right now is not enough. And it's going to go on being Not Enough for quite some time..." Let's also focus on Exp. Tools! www.science.org/content/blog...
Looking for a Postdoc position in #Genomics at the interface of computational and molecular biology? -> The Rube lab at UC Merced is recruiting. I highly recommend you reach out! aprecruit.ucmerced.edu/JPF01874
Thank you! Very complementary to the synthetic platform your lab built. Still challenging to go beyond 5kb for us, so tackling distal enhancer regulation will take a lot more engineering! Looking forward to future work from your lab!
Scalability is definitely the next big step in the journey. Also in the works a universal replacement cassette. With the current throughput it can still be useful for benchmarking/providing insights into potential shortcoming of "AI" tools.
Thank you!
