Kamal

:)

So we’re about halfway through the life of our sun. So maybe 4 eons left? Like the 4 before us.

Belome is a 4-eyed, 4-horned, bipedal people eater with a canid shout, a lolling tongue and sabered teeth. They have red spiral markings in lieu of spots upon a tallow-toned coat. They are remembered as an original character and final encounter in the Kero Sewers submap of Super Mario RPG: Legend of the Seven Stars. They are weak to lightning-element attacks and to their own perpetual hunger.

Looks like Belome from Mario RPG

“smolts” :)

Even us TAs were caught in the blast wave

This is a very elegant explanation—thank you!

Will do ✨

This is super cool! Are you using RFDiffusion/ESM as a starting point? 🤔 and then adj and ss are precomputed with other tools.. so they need to be in .pt format to be ingested by the model? I haven’t tweaked model inputs at that level before—I’ll give it a think and ask around~

I recently learned these are carnivorous (they eat slugs and bugs). Does anyone know how to raise carnivorous larvae? For repopulation purposes… they’re going extinct here.

Is it possible to insert adversarial content for tampered browsers and headless traffic? Like replace the text with random letters so it’s untrainable, or render nothing. Also if the user doesn’t consent. I guess you would have to mask for the firehose as well? That could make it harder to federate?

I’d forgotten the name of Polari the other week, and none of my rhymed memory fragments were autocorrecting the right way when I tried looking it up. Thanks for leaving a breadcrumb!

✨

Super cool! How large of cargo can you put in (say, an entire cas9 + template)?

That’s interesting! I struggle with tons of false positives in protein-protein AlphaFold-type simulations—since it seems to clobber its way through the interaction on the assumption it does interact. Maybe ligands have a false negative problem? I don’t have an intuition for it.

Slides from a presentation describing a hearsay history of the R tidyverse package.

I just went over an abbreviated version of the history in the Biostats class I TA. Is this accurate? I feel like it’s a bit hearsay from what I could find. I’m looking forward to your authoritative version! To correct what I got wrong :)

I was surprised by @martinsteinegger.bsky.social's recent paper stating 99% of viral proteins to be <1500aa long, since I've never thought about it before. This cropped figure from Yannis Nevers' analysis last year doi.org/10.1186/s130... suggests all domains might similarly have 99% < 1500aa! Cool!

I actually wanted to use twitter as a comments engine in the past but couldn’t get the whole conversation to show up with the supported embedding setup—great to see it’s actually possible with 🦋

This is super cool! The structural diversity is very exciting ✨

If the rest is of interest, I’m the molecular plant-pathogen interaction guy!

👋 Thanks for posting! I just applied -- I didn't tailor my research statement to healthcare though, I just loosely described one of my thesis aims. I hope it's sufficient!

Forgive the mediocre resolution -- is this the full text?

Is this because any reversible/lossless compression has to fully represent the initial domain to be reversible or does the network learning on it have to learn how to unzip? I guess it would learn some representation of the (un)encoded input regardless 🤔

Terra preta - Wikipedia

I’ve been a terra preta (en.m.wikipedia.org/wiki/Terra_p...) fan myself—dunno if it made it up to the Mayans in the Yucatán!

This reddit post is pretty informative: www.reddit.com/r/bioinforma... the part that jumped out was point 2: “ the absence of protein-internal solvent atoms can easily result in collapse or destabilization of the protein's structure” so you need to add water atoms and stuff.