Thanks to you and the team! Can I ask: I find I have to zip bruker.d raw files (which to be fair look like folders), the tool thinks they are folders and it tries to upload the contents as separate files. Is there a work around to this? I did check the troubleshooting docs. A while ago admittedly!

I suppose this correlation shouldn't be too surprising... I'm always thinking about how dynamic the transcriptome is. To me it's not massively surprising it doesn't correlate all that well with the proteome for this reason alone. Nevermind all the proteasome buffering, translational magic etc...

Our mixed species experiments (aka. CQE, LFQ bench) show excluding precursors below any PPP thresholds can make protein quant objectively worse (and cuts IDs). In short-using more precursors is better than excluding some. Average PPP overall might be useful though- I wouldn't want this to be low.

*Ignorant blueskyer* why do we hate RG now? Is it a colourblindness issue? What to do instead?

It's a problem with people not adding it for sure. Most people use some sort of alkylation but don't always add mods they aren't explicitly studying.

This put me down a small rabbit hole which led me to the quantms-utils github, where I found this gem "diann2mztab". Am I right in thinking this will allow complete depositions of DIANN results with proteomeXchange tool? Do you happen to know? :)

P.s. poster F42

Don't know these guys but a "$50 proteome" when you read the press release is actually 1000 targeted proteins. I hope decision makers know that this probably means they are missing things...

Turns out Humpty Doo is a place. Which is fair enough!

We see this sort of thing all the time... our general rule with this is "if you're comparing conditions just assume the most extreme ones on the volcano plot are almost certainly correct, be cautious about the others".

Yes, of course a predicted library is still a library 😅 especially after MBR... yes, then that is crazy fast.

If one says "our LC-MS is amazing compared to others", one should be 100 % certain they are doing everything correctly. Your point on double normalisation is also well noted, I think people do often use normalisation with software followed by median normalisation. Is this good practice?

Library free??? If so WOW

That's great-thanks for showing this, will definitely keep an eye on the m/z delta, I noticed there are a lot more plots in the QC report pdf, will definitely check this if we get low IDs! Thanks to you and your team for the hard work on this amazing community resource 🙏

I have some colleagues who use the Orbis more than me, I will mention this.

Likewise, EuPA is very much on our radar. Its a shame, we love going to ASMS when we can.