Saturday, December 17, 2016

Is this new paper showing we could have 4x more Orbitrap resolution!?!?


I'll admit it, the title of this paper sounds pretty boring.

It painstakingly details a new way of processing Orbitrap data that these researchers call ΦSDM.  This is definitely a development paper because it shows that this algorithm has some downsides that will have to be addressed, but it shows a crazy amount of potential as well!

Downsides?

Computationally far more intensive than the normal algorithm(s?) running behind the scenes in your instrument.

May be negatively affected by having more stuff in the Orbitrap -- actually -- that is how I read early into the results, but later on into the paper (and a lot more math I don't understand) this actually starts to look like they work their way around it by "binning" frequency. I'm still going to leave it as a maybe negative.


Potential?!?!   Is this 40,000 resolution at 32ms transient!??!?


The 32ms transient on a QE HF in ones I've used should get you somewhere around 15,000 resolution around 200 (m/z).

Since Orbitrap analyzer resolution is m/z dependent this is what we'd normally see on a QE HF running at 15,000 resolution (in this case, I just grabbed a quick MS/MS spectra; click to expand)


Around 200 (m/z) in a normal "15,000 resolution" Xcalibur reports around 17,500 resolution for an ion near 200 and 9,000 resolution for an ion that falls in the range of the SDM image above for the 32ms transient.

40,000 resolution is 4.4 times more resolution!!  If we just think this might be linear -- this is something in the range of 75,000 or 80,000 resolution with a 32 ms transient!!!!  (Sorry, I'm excited!)

Okay...I'm just going to extrapolate just a bit here. And I'd like to lead with the statement that that this is just me going into my crazy mass spectrometry fanboy thing while we are very snowed in on a Saturday afternoon -- what if this was linear? What would that mean for us?

In term of getting lots of resolution -- and this is linear -- our new transient vs resolution chart for a high field Orbitrap might look possibly maybe look something like this --


RIGHT!?!?  Again -- I don't know what many of the figures in this paper mean (I'm a biologist, remember?) but just imagining the possibilities of this makes me very very happy.

From a proteomics perspective, this is no way the most exciting chart. The most exciting chart would be going the other direction. I really don't need 80,000 resolution in every one of my MS/MS scans (holy cow -- could you imagine how big those data files would be?!?!  Would that be 10GB/hour coming off each instrument!?!?).

The more exciting thought is this --> how fast could I generate 15,000 resolution scans if I had enough computational power to run this algorithm instead of the normal FT algorithm? Could we get 50Hz in an Orbitrap? Sure is a nice thought!


Shoutout to Mass Spec Pro for breaking this one on Twitter and giving me something fun to read this afternoon!

No comments:

Post a Comment