John Trant

I have found that posting inner thoughts and responses to my wife @shufflersunite.bsky.social on Bluesky, although entertaining for @sculpturatus.bsky.social and many others not as vocal, is not, in fact, the same as keeping those thoughts to myself.

Go figure.

I KNOW RIGHT‽‽‽‽‽‽‽‽

Clearly an error. Screenshot of Google scholar with a paper entitled "Please do not adjust margins" likely associated with a recent paper from our group on the Tn antigen published in the New Journal of Chemistry.

According to Google Scholar, a new paper from our group just dropped.

Sounds like you needed an exorcism.

I can't believe you are making me agree with a biologist.

HVac varies greatly in its ability to remove all solvent.

Its why we label with a pencil on a paper sticky label, taped over onto the vial. Or our newest ones use a (solvent resistant) barcode and ink on a printed sticky label.

Never again. Jenn, you were an example we all learned from.

A brilliant colleague of mine in my PhD was fnihsing her degree. Had all her vials out to do her final characterizations. Undergrad spilled a bottle of acetone over them. They were sealed...but all labels were in sharpie.

Now had a ton of vials but no idea what is in them.

Yeah. But, not sure I agree on the later. How much money has OpenAI raised on the same product? Like, good quality hype should go for a few orders of magnitude more than that.

Research ethics question:

Do i need REB approval to use human blood i collect from someone who is hemorrhaging for research? I am sure the answer is yes. Someone else is telling me if I call it "pilot data" I don't.

A-game emojiing

#We'reATeam !
#GoSensGo

Sorry to my Habs-loving colleagues. You are good people.

No.

I think proposing to ban all muslims from the US or certain positions demonstrates a clear statement of religious bias.

This is true. I wonder what the global user numbers who skeet primarily or partially in English are? I bet it leans a lot more less-US centric than X.

Bluesky seems pretty centrist to me. It's not really left leaning at all.

The fact that the political middle ground is seen as "left leaning" is really weird.

Just got a LinkedIn notification for a former mentee of mine on his 4 year anniversary at his job. So he's got a PhD. I mentored him as an early degree undergrad when I postdoced with the incredible Beth Gillies @westernchemistry.bsky.social in 2012. I SHOULD be publishing that paper in 2026...?

This is the entire package of data, and our insights, such as they are (things are always complicated so any INSIGHT should always be viewed with maximal skepticism) are all there.

Plus, the pictures are pretty pretty.

So here is a few hundred hours of highly skilled postdoc time thinking deeply about crystal structures with some computational analysis tossed in (its a review...so no new data...) for free. Use it. Contribute. And then cite us please. I promise we aren't holding back something juicy for our own use

The main contributors to this field have done a ton, but their reviews don't really drill down to this form of comparative mechanism across the family of known compounds. We are new to the field, and so we wanted to know where to start. And we found that the analysis was missing.

Anyways, you can go download 3D images and manipulate them and see how all the work everyone has done in this field contributes to a good baseline for discovering new members in this class. We developed this tool for our team, but there is a lot to do.

Like scientific articles are the things that should never be summarized by AI. Ever. It's like saying "here is a Shakespearian sonnet. Would you like me to summarize it for you?" or "here is a love letter from your future partner at age 14, would you like me to summarize it for you?"

tRANT over.

But the part you want to read? You should probably read it. All of it. Because we kept it as short as we could while saying what needed to be said. If more could have been cut without cutting content, we would have.

Any errors are ours. Any bloat, I blame on the reviewers.

Any text that isn't necessary isn't there. If you summarize you either get:

a) A stupid version of the abstract.
b) A random selection of random facts.
c) A likely misrepresentation.

You might not want to read all of it. Cool. I wouldn't either.

An abomination. On the left is a table of contents well organized for your reading pleasure. In the middle is the first page of a pretty good paper about carbonic anhydrase. It is long. Like 20,000 words long. With like 20 multipanel figures making complex arguments about the finest details of molecular interactions and spatial relationships. On the right is an offer from the publisher to summarize it using AI. This is what one sees as the author when told "congratulations for publishing your paper!" and something about the value to the community. And then the message from the publisher is along the lines of "how about we ignore the value of the carefully considered and revised text edited by 7 authors in a years-long iterative cycle, and just get a bullshit machine to spew some spew from it, I'm sure the other stuff doesn't matter." I hate this commodification of research. It's like the content of the paper isn't important. The only value is the existence of the paper. If I wanted lots of papers for the sake of lots of papers, I'd be an inorganic chemist (no offense to my "toss a ligand on the metal, change the metal, change the ligand, get different papers each time instead of just throwing 20 papers worth of stuff into a single SI so you can make an actual scientific argument in a single paper" colleagues). I'm probably going to regret this last paragraph. But I need the laugh.

You open it up, and the first thing that hits you is an option to summarize the paper using AI.

No.

You know what? There is an abstract. I curated that. It's called a summary. The rest? There is not extra text. That got deleted. There is enough text to explain.

tRANT time.

A new paper getting published is exciting and all (not a first paper for anyone here, but it is a nice piece of work if you like carbonic anhydrase inhibitors and want to think about design features for isoform selectivity). doi.org/10.1016/j.ab...

BUT...(next panel please)

If I ever accept investment, they'll probably make me change my business strategy and hold me accountable for my donations and stuff. So, holding off on taking people's money on this for as long as I can so I don't need to justify anything to anyone but the Tax Man.

We are lean, so no HR and no employees. If it ever makes sense, then we want it to make sense. But BSRS isn't about making money. It's about solving problems.

So this also means that I can both be an anti-capitalist market socialist and run a company with a clear conscience.

I do note that this means that I earn -infinity ratio compared to my lowest paid contractor/employee (my only contractor, my CFO). Take that Tech CEOs! (not that they see any problem with paying people minimum wage and drawing trillion-dollar paypackets themselves).

We are for profit as the taxes are easier that way and it gives us a buffer if something goes wrong and I need legal or other support. So it makes sense. But it isn't a traditional "take all the money you can" company. Not how I roll. Will never be how I roll.