Thrilled to share our paper out today. I am immensely grateful to my mentor, Dr. Ömer H. Yilmaz @ohyilmaz , and our fantastic collaborators @JudithAgudo1 & labmates and @DamonRunyon for five years of support.

the bureaucrats buying bunsen burners fear the Lab Candle Enjoyer

Great lab bonding at our 2025 lab retreat. A good mix of scientific discussion and lab fun. I'm so grateful for such an inspiring and energetic lab!

being the first to jaywalk is so aura inducing the other pedestrians all look at each other like damn he’s really him

Happy to announce that I was awarded a pathway to independence award (K99) from the NCI/NIH today to work on studying whole-body metabolism driving cancer cachexia. None of this would be possible without support from @DamonRunyon and mentorship from @Transactivator and mvhlab.

Uh oh: deep learning models in genomics lose badly to very simple linear models. Could they have been over-hyped? (new paper by @s_anders_m @const_ae and Wolfgang Huber)

calenderheads peep february 2026. you dont see em laid out like this often anymore. something to look forward to

‼️Excited to share my first preprint in the Stegmaier lab – using metabolite depletion screens in physiological culture medium and in vitro/in vivo CRISPR screens, we identify metabolism of vitamin B2 (riboflavin) as a potent dependency in acute myeloid leukemia (AML). A 🧵…

If I'm reading the new STATE ( Adduri et al. , https://t.co/8MfbP1ATJb ) paper correctly, we still cannot predict the effect of genetic perturbations better than a simple mean, and many methods do worse?

AHH!! My primary thesis work is finally out! Tweetorial to come...

im not a big gossip but i love Knowing Things. so when i ask for gossip it's never sinister. it's pure curiosity and a desire for knowledge

Recent Beamish award winner @TaoThuan @NatMetabolism @UBC describes how to avoid “nonsensical” results in metabolomics https://t.co/mtOMxHb74p

A case for substituting butter with plant oils From >220,000 participants in 3 large cohorts using validated food questionnaires with 33 years of follow-up https://t.co/1ReOBeobXn

Honoured to have been invited to this symposium! Had a great time presenting my work and interacting with other amazing students!

will forever point out that PhyloP, a traditional probabilistic model, has a fraction of the compute cost and does as well or better than many of these foundation models

Seasonal depression “isn’t real” until you get a warm day in February and realize you’ve been dead inside for months

https://t.co/GGP3kE9Lk1

I can’t believe it - after years of advocacy, exclusionary zoning has ended in Cambridge. We just passed the single most comprehensive rezoning in the US—legalizing multifamily housing up to 6 stories citywide in a Paris style Here’s the details 🧵

New preprint claims that most existing DNA language models perform just as well with random weights, suggesting that pretraining does nothing (Mistral & DNABERT-2 look like exceptions). We need better DNA language models.

and I don't just mean papers - I mean reagents, code, IP - everything