Funny quote from a conversation with a friend at lunch who accidentally took an interview from a biotech company founded by a 20 year old Thiel fellow: "Silicon valley didn't learn from Elizabeth Holmes, they actively doubled down."

“Inaccurate energetics are evidenced by the inability of AlphaFold3 sampling to reproduce experimentally derived ensembles. Physics-based evaluation identified system-wide limitations in AlphaFold’s ability to make physically accurate predictions.” biorxiv.org/content/10.110…

x.com/keyonV/status/…

Practically useful & biologically aligned benchmarks such as this one from Peter Koo lab consistently show that all the overhyped annotation-agnostic DNA language models are actually terrible for transcriptional regulatory DNA in humans (mammals). 1/

Prof. Nikolai Slavov This is a really interesting study. It would be really interesting to try estimating protein translation and protein degradation rates. Have you thought about this?

Josh Welch Andrew Andrew Leduc talked about measuring protein synthesis and degradation rates in single cells at Single-Cell Proteomics Conference: youtu.be/adkY6txDyqs?si… We are working on writing up these results soon, and will be happy to share the data with you.

For biological data, if you don't have deep expertise in this low value work called data cleaning, u r lacking a fundamental understanding of the idiosyncrasies of the data. Without this knowledge, it is impossible to seriously model data.

Ryan Morey Possibly. I wanted to simply clarify that in compbio the fundamental problem is often the opposite - that early researchers don't spend enough time wrangling, understanding & sanitizing raw data

We traveled to DC to talk to Congressional Staffers about the impact of NIH NIBIB funding. It enables us to develop synthetic biology tools to advance cell & genetic medicines for patients in need. REPOST to deliver the message that SCIENCE saves lives & makes America great!

It's that time again, where finishing the last 5% of a paper means your productivity turns to snails pace

Many proteins bind RNA, yet we still don’t know what RNAs most bind because methods map one RBP at a time. In Cell, with the Jovanovic lab, we describe SPIDR – a method for mapping the RNA binding sites of dozens of RBPs in a single experiment. sciencedirect.com/science/articl…

One lesson of DOGE is that a lot of federal govt spending is sent to old ppl (social security, Medicare) and affluent individuals/orgs (defense, health spending) but the spending with the highest “lives saved per dollar spent” tends to go to v poor people or to scientists working

In this interview, I explain how university research, when funded by independent #NIH #NSF federal grants, brings oversight to biotech to fuel innovation while keeping science transparent, reliable, and in service of the public good: hub.jhu.edu/2025/07/30/jea… #ResearchMatters

The NIH budget would INCREASE by $400m — not slashed by 40%, as Trump proposed — in a proposal by Senate appropriators, says Sen Murray. "Some have asked if there will even be an NIH by [2029]. The commmittee's resounding message is yes—Congress has your back", she says.

people arent gonna wanna hear this but i truly do not believe this mistake could’ve been made by someone with a phd. after going through brutal peer review several times you just stop doing stuff like this. whoever made this graph clearly has a bachelors degree. maybe a masters

This dude is insufferable

Never cease to be funny that hypers & grifters confuse or pretend that technological capabilities that could be robustly implemented in 30 years can be successfully deployed today.

Haven't been in the lab in a while, forgot how fun it can be :)

"Tabular data" didn't make sense to me as a category. Images or text are generated in a pretty specific way, but you can put anything in a table. These points about common properties have helped me understand why the same methods might work across different tabular datasets.

🔥