Vinny Costa

Josh Siegle and I are thrilled to be chairing the first workshop on "Bridging the gap between cell types and spike trains"! We see this as the key link between population-level descriptions of dynamics and real mechanistic understanding from cell types.
celltypestospikesworkshop.github.io/2026/

Ambient RNA & barcode swapping is a serious issue in single-cell genomics. Tools such as CellBender, scAR, DecontX & SoupX. We have developed CellSweep which is faster (in some cases by a lot) and much more accurate. Extensively tested and benchmarked. www.biorxiv.org/content/10.6... 1/

When I read your post, all I could see was this

“Area 10m is the only common ‘driver’ of both pgACC and sgACC”

New connectomic scale understanding of ACC from Daulton Myers and Julie Fudge. #mustread #neuroskyence

www.biorxiv.org/content/10.6...

Ensure+methamphetamine.

Nosepokes go to 11.

To be clear, I think this is a great demonstration of the difficulties in equating reinforcers across task contexts and stimulus modalities that confounds a lot of valence based neuroscience research. It’s a hard problem.

…but the footshock elicits a much much larger “dopamine” response and that does scale with US magnitude. Doesn’t the BLA flight response just reflect US valence (or saliency; the weak appetitive US confounds interpretation)?

I am curious about how you are interpreting Figure 2.

If the sucrose and Ensure are just overall weaker reinforcers relative to the footshock, and the BLA “dopamine” (genetically encoded sensor response) doesn’t differ between them….

T. rex took 40 years to reach full size, study finds Tyrannosaurus rex may have taken far longer to grow up than scientists once thought. By analyzing growth rings in fossilized leg bones from 17 tyrannosaur specimens and using new statistical methods, ...

That's because they coddle them too much now. Back in my day, it only took 20 years for T-Rexes to fully mature, because they had discipline. Spare the rod, spoil the terrible tyrant lizard.

www.sciencedaily.com/releases/202...

Jealous

Also Echo’s customer support is very responsive and no bullshit. I look at one PDF on the Keyence website and 47 different reps call me to sell me a rock tumbler.

I get that there are fancier better scopes, like those from Keyence or Zeiss. But as an experimentalist I want shit that works on a mesoscale level and doesn’t require a PhD optics to use. Echo’s scopes fit the bill 100% in that regard.

Give me my images and let me get on with my science.

I have a Revolve in the wet lab that literally has revolutionized how we do immunostaining.

And I bought a Revolution stitch large format Z-stacks.

I am salivating over there scanning confocal.

On the Echo, I take an image, I transfer it, and I just plop it into ImageJ or QuPath and start analyzing. No 5000 clicks and buttons and options and other fuckery to slow me down.

All microscopes suck, some are useful.

Slower acquisition than the Echo Revolution. Also the Analyzer software is terrible whereas the Revolution automatically computes the EDF on the fly and just generates a stitched image at the for direct OME TIFF output. I hate the Analyzer software. It is completely non-intuitive.

@tollkuhn.bsky.social

I still found BZX-1000 was slower than Echo Revolution for stitching large format multichannel Z-stacks for macaque tissue. Echo scope was also substantially cheaper and I have undergraduates using it on the regular. Bonus the software is far more intuitive.

Congrats!

Thrilled that this first empirical paper out of the lab is posted, led by Sandarsh Pandey, asking:

Depression (and other internalizing disorders) involve profound changes to sense of self. How can we study these differences using rigorous decision-making methods?

(alt link: tinyurl.com/2kk59dje)

Graph of award probability of R35 and R01 from NIH factbook as a function of review rank percentile. As is apparent, 2025 is a significant departure, with lower award probabilities at all scores <40 and significant departures from norm, where even being in the top 10% is no longer a nearly certain indicator of success. Data source: https://report.nih.gov/nihdatabook/report/302

The data is in: the NIH goalposts have shifted.

What were once almost certain fundable scores have become coin flips and what used to be likely grants have become aspirational, leading to fewer awards.

Another manifestation of how HHS policies have led to fewer awards and less science.

Optical illusion of a woman bent over some papers. Her sunglasses are pushed up and she is wearing a hair band so the top of her head looks exactly like a Muppet face

Sorry I know the world is in a terrible fix but I've been laughing at this for ten minutes now

We made a highly blue-light resistant red-fluorescent genetically encoded calcium sensor GECI) www.biorxiv.org/content/10.6...

BluePrint™ 🗺️ below:

Great except the wiring of the brain isn’t especially random (c.f. the fornix).

Raccoons optimally forage for information: exploration–exploitation trade-offs in innovation Animals in rapidly changing environments, such as cities, should optimize information gathering to learn about and efficiently forage in these heterog…

Trash pandas show effort-based discounting of “exploratory” decision making.

Optimally claims aside, I genuinely love this study.

www.sciencedirect.com/science/arti...

Bring On Defunct: The iPod Enthralls Young Music Listeners

Happy to sell my poop brown Zune to the highest bidder…

Bring On Defunct: The iPod Enthralls Young Music Listeners www.nytimes.com/2026/03/01/t...

Bellair’s? Jealous you get to check that off the bucket list.

Reach out, happy to help get your lab started using the RMAs.

Should work in marmosets if Fc receptor genetic overlap is high enough. Actually given smaller brain volume & higher neuronal density/mm I bet you would get even higher SNR, similar (rhesus brain is too damn big). But probably best to optimize to each species.

Thanks Eliza!

This. is. amazing.