I'm sure @leecantrell.bsky.social will do a real tweetorial, but my favorite: "Reduced peak completeness minimally affected detectability (LOD) but substantially degraded quantifiability (LOQ), indicating that identification-centric readouts can remain stable while quantitative utility declines."

They're calling it the stupidest thing you can do in an era made almost entirely out of stupid things you can do

This cartoon hits hard 😢 #JimCrow2.0 via @seattletimes.com

The most accurate portrayal of the pro-Jim Crow, white nationalist Roberts Supreme Court I've seen yet.

My partner was just asked for triple verification at the bank. Offical government ID alone wasn't cutting it. We're in a new and frigentening era where only through corporate proof that we exist that we can prove who we are.

It is hard to argue with these qualifications. www.cnn.com/2026/05/15/p...

#proteomics_101 Provide an error analysis that demonstrates how much additional confidence should be given to a peptide sequence match made by multiple algorithms as compared to another assignment made by only one algorithm.

All the car parks to the supermarkets and hyper markets in this area are done. Plus quite a few much smaller parking areas. All new agricultural buildings have solar panel roofs as well.

CIHR has launched a trial ballon that suggests shifting to one competition a year. This is the worst idea I have heard in a long time. The only benefit is making top-line success rates look better than they are. Keep it as is, do in-person reviews, and keep the pain in the open.

when I googled “Dumpster on Wheels”

#proteomics_101 How does the N-degron pathway affect the N-terminal residues possible for type I or type II membrane proteins? Explain why it does not affect N-terminii generated by mitochondrial targeting peptide removal.

This cartoon is sixteen years old but has not aged a day.

Conestoga did plenty wrong. But most of it was approved by the provincial government on a nod and a wink. For the province to come in and put the college under administration now, on the basis of an audit it has not publicly released, is some serious bullshit.

#proteomics_101 In yeast, ~10% of proteins are classified as "essential". Would you expect that percentage to be higher or lower for proteins involved in large molecular structures, e.g., the large & small ribosomal subunits?

"Based on the dataset you shared, US and UK responses differ mainly in tone, intensity, and wording style, even though they express similar emotional states" New post on what happened when I got Copilot to do some data analysis: kucharski.substack.com/p/real-signa...

Thanks to the Canvas debacle, I am now getting an exciting new genre of spam from entrepreneurial undergrads who want my institution to license their alternative LMS startup offerings.

#proteomics_101 Yeast COX15:p has a predicted mitochondrial transit peptide (1-33) leading to N34 being the N-terminus. Proteomics observations indicate the transit peptide is (1-66) leaving P67 as the N-terminus. How would you investigate this difference & make your findings available to others?

#Proteomics desperately needs a meeting without any vendor presence. No product seminars, no new instruments, no new methods. No fucking AI. No new tech shit at all. Call it "Back to Biology" and try to remember how to do science. Get off of the hamster wheel, for God's sake.

LLM weirdness: I have noticed that very recently search engines have started to substitute gene names without any apparent reason. For example, this morning I searched "ATO3 mitochondria" and many of the answers came back with ATAD3A (an unrelated protein) articles instead.

Cdn universities and colleges are collectively going to pay well into seven figures in ransom over this. Maybe even eight, not sure. You would not believe what data some people store on these platforms. Also, I've learned that there is such a job as a "ransom coach" and they're in very high demand.

Of course, the past quarter-century+ of policy in the Anglophone OECD has strongly implied if not explicitly claimed the idea that attracting bums to seats via market-like competition was the best indicator of quality.

When proposing what you consider to be a breakthrough in understanding of a protein's function, you must be prepared for the fact that you are simply adding yet one more function to what may already be a long list. It is way the literature/databases work & there is nothing to be done about it.

At some point, repetition ceases to be an homage. We are passed that point with "Colossus". en.wikipedia.org/wiki/Colossu... en.wikipedia.org/wiki/Colossu... en.wikipedia.org/wiki/Colossu...

Oh boy! Another consortium! I am sure it won't be just an announcement paper & not much else. Even though all previous consortia in the field (at least 1 per year since the mid-'00s) have followed this pattern. www.nature.com/articles/s41...

This is the kind of doctrine you can only fully embrace after sustaining the kind of highly specific neurological injury that is induced by taking an economics degree, an injury that makes you incapable of perceiving or reasoning about power. 7/

A revelation that came to me today while watching 2 guys trying to do something I found inexplicable was that I may not be dumb enough to comprehend some human behaviours.

Finally, a major American newspaper whose editorial board is courageous enough to stand up for the rights of big tech companies.

Canadian weather forecasts to become useless soon! www.ctvnews.ca/climate-and-...

Not loving this weather pattern (wind @ 250 hPa).

Global aluminum price: $3,500/t U.S. price, with tariffs: $6,100/t Trump's dream of a purely domestic auto industry is total fantasy, but industry is enabling it because they think he'll find some scuzzy way to make them whole. Good luck with that.

Some from other countries may flee to Mexico.

Pioneering scientist J. Craig Venter has died at 79. His "whole genome shotgun method" helped genome sequencing become faster and cheaper. n.pr/4eSO7cc

#proteomics_101 List the 3 most commonly used experimental methods to determine if a selected region of a protein sequence is a transmembrane domain.

One of my least favourite styles of book is a bunch of essays on related topics selected by an editor that have been stapled together & pretend to be the "bible" for a difficult to define subject.

Having grown up on a farm with a 2 acre vegetable garden, I find growing microgreens to be an odd but satisfying hobby.

I've dropped Visual Studio Code today. I don't know who's at charge of these things at @microsoft.com, but they seem to be completely disconnected from reality at this point or care that they're breaking what's left of our trust. github.com/microsoft/vs...

I love the formulation that LLMs generate "artificial information": looks and feels like genuine information but isn't.

Polybromo 1 (PBRM1:p) has nothing to do with bromine, but its 7 rigid nward bromodomains have affinity for ε-acetyl-lysine. The pS/T & pY patterns make it clear that the each rigid domains is not regulated separately, with only one of the dynamic IDRs taking on most of that burden.

If your business has a website, you may want to have an app developed. Here’s a quick guide on how to make that decision: Does your website function? If the answer is yes, don’t release an app. If the answer is no, fix your website and don’t release an app.

I love these things. Cheap & solves all sorts of solar lighting issues without me having to make anything.

#proteomics_101 In the human proteome, the ratio of S:T is 1.6:1. In a phosphoproteomics study, it is not unusual for the ratio of observed pS:pT to be 6:1. Explain this difference, with specific reference to the idea of lumping together all serine and/or threonine kinases into a group.

I'm really enjoying imagining a scientist who's driven by hatred of isoleucine

Very sad news. Venter was one of the most impressive scientists I met during my studies (or anytime since).