Over the last few years I think I've seen 5 different company presentations that have been in some sort of an arms race - purely with one another - to deep dive in human plasma. This one now says they can get 3k proteins in human plasma, so the other one says 3,500 and the next one has to say 4,000, and then 5,000. I've visited a couple of them personally and the scientists running the samples seem to spending most of their time trying to find new jobs because their executives and their sales teams seem incapable of shutting the fuck up for even 10 seconds and are pulling these protein coverage numbers out of their own lower body cavities. Yo, if you're ever in a company as a scientist and you answer to the sales team - get out. 12 years hanging around mass spec companies and I've never once seen that setup work.
Back to the paper!
Among all this noise there is have been credible developments in hardware and chromatography and nanoparticles. And what if you really sat back and had people who are experts at all of these things work together to see what they could really do?
The best case scenario is that you might end up with something like this.
This is the newest (that I'm aware of) generation of Seer Proteograph linked to some fancy uPAC 50cm chromatography running at 1uL/min on the Orbitrap Astral with about 30 minutes/sample and a similar but slightly longer method on an Orbitrap Exploris 480. 400ng of peptide (wow! in my mind now, that's a lot, but people put a lot) on the Astral and something in that range on the Exploris. Feels like they tuned in the chromatography on the Exploris and then the Astral arrived and they were good to go on the chromatography side.
Of course, the nanoparticle corona stuff had to be done up front and that's where the magic is to get past those 22 pesky proteins that make up 99% of the blood plasma.
Now - this isn't a SomaScam /Illumina Protein Crap experiment where you can detect all sorts of stuff but you can't quantify any of it because your aptamers have a dynamic range of 2 (not orders, 2). This group spiked in bovine proteins at different levels in different samples and determined how well they could quantitative recapitulate the expected ratios.
Turns out it works really really well, with or without depletion. For real, there were legit analytical chemists doing legit analytical chemistry here and it makes this whole workflow seem very very smart.
Sure - this isn't an inexpensive workflow. We all know everyone complains about the Seer kits. I hear rumors of $250-$350/sample all the time. And an Astral isn't cheap, but we're talking about 48 samples/day so 336 samples/week before controls? That's more than Illumina's solution can do. And - sure - an Astral isn't cheap but neither is an Illumina sequencer (which you also need for O-Link)
It's been a pretty good week for plasma proteomics for mass spectrometry.... and sure, I'm biased, but if we're competitive from a price, thoughput and coverage perspective, the other technologies seem more than a bit silly.
