Thursday, October 10, 2019

Scheduling PRMs for lots of metabolites and/or phosphopeptides!

I've spent a lot of time with this great paper this week, and even though I posted it on another blog a long time ago I just realized yesterday how useful it is for scheduling PRMs for proteomics.

Here is the scenario -- you discover all this cool stuff with your super cool ultra high resolution system -- now you need to verify that this doesn't just occur in your 20 patients from your concatenated highly fractionated pool -- this repeats in your 300 person cohort. If you've got a super fast triple quadrupole system (or "double quad" [if you're using a Perkin Elmer [please DON'T use a their triple until they become 100% Skyline compatible! Don't let my mistake(s) be your mistakes!]) maybe you don't care. If you are pulling 600 scans a second, just copy your whole list over to it and don't schedule a thing. Set up a 10 minute run and walk away.

However -- the fastest Orbitrap is going to knock out 40 scans/second. That's a lot less targets and if you're targeting you probably want to allow a decent amount of fill time -- the Orbi still can't get as much signal in 10ms as a triple quad can with 10ms of dwell time.

In this study the authors demonstrate an easy way of generally timing PRMs and it works better than you'd believe.

Metabolomics length runs (sub 30 minutes)
PRM targeting on a D30 (Q Exactive Classic, so maximum 13 scans/second!)
237 metabolites quantified!
The scheduling is worked out more along the lines of segments rather than ultra tight scheduling. I think I told someone this week they did "early" "middle" and "late" eluters. A quick reread suggests that they broke these into closer to 7 segments based on m/z.

I think this provides a great idea of the upper limit of what is possible when targeting with PRM on an Orbitrap instrument.

While on this topic, I would like to mention another resource that should be getting more attention for this sort of thing -- focused on phosphorylation site PRMs!

At the Phosphopedia you just find or build your pathway and it builds your PRM assays. If something exists for other targets and I don't know about please email me!

Here I just chose the AKT phospho pathway -- the whole thing -- and it makes a list of my PRMs and I give it the +/- retention time wiggle that I think my nLC system has (+/- 5 min on a 60 min gradient)

-- and here is my targeted list (import directly into your QE as is) but here is also the schedule density. Around 10 minutes I'm looking at 20 targets. If I'm using a QE Classic and matching my fill time to my transient (roughly 50ms fill time with about 14ms of overhead) 64ms x 20 = I'm only going to get a PRM on each individual target ever 1.2 seconds. Just fine if I've got a 30 second peak width --I could probably get away with adding a few more targets at the 10 minute time point -- not fine if I'm doing UHPLC and my peak width is 4 seconds. I'll either have to work on narrowing that retention time, or drop some targets out of this run.

1 comment:

  1. I suggest you check this out: