is the *order* of nearly isobaric centroided peaks in an orbitrap consistent from scan to scan?

[Gautam Saxena]

In an orbitrap (specifically a Fusion or QE running at 30k resolution), let's say one has successive scans that have been centroided using the built-in centroiding ability from the newest Thermo .raw file reader sdk. Let's say further that there are several centroided peaks in each scan that are very close to each other (ie within 1 or 2 ppm). And let's say that we know (through observations) that each peak is accurate to say 5 ppm. (So, in this hypothetical example, the peaks in each scan are *closer* to each other, ie within 1 or 2 ppm, than the overall observed accuracy of our orbitrap at 30k, eg 5 ppm.) Then, is it always true (or 99% of the time true) that while the accuracy of any peak is only as good as 5 ppm, the *order* in which the centroided peaks appear in successive scans is the same. 

For example let's say at scan 100 we have a peak at 500.00Da and 500.01Da and in scan 101 we have peaks at 500.01Da and 500.02Da. And let's pretend that there are only 2 analytes in our made-up sample. And, let's say analyte #1 caused the peak at 500.00 Da in scan 100 (and therefore analyte #2 caused the peak at 500.01Da in scan 100.) Would it be always true to say that in the next scan (eg 101), the 500.00 Da must correspond (or >99% chances are that it corresponds) to analyte #1? (And let's just say, for argument's sake, that we have observed that the orbitrap's accuracy at ~500.00Da is approximately +/- 0.03 Da in this case.)

(Or, does this situation never really happen, since we can always assume that the difference between centroided peaks within a scan (ie due to the instrument's resolution) is always larger (in fact at least twice as large) than the difference between successive scan's observed mz values. (I believe, but I'm not 100% sure yet, that this situation does in fact happen, so I believe the question posed above is valid, but again, I'm not 100% sure....))

[Gautam Saxena]

Quick correction to my example. I should have said: 

Would it be always true to say that in the next scan (eg 101), the 500.01 Da must correspond (or >99% chances are that it corresponds) to analyte #1?

[David Bouyssié]

Dear Gautam,

It's difficult to answer your question because it totally depends on your full setup.
For instance, are your coupling your instrument to a Liquid Chromatography separation system?
If so, then new analytes may appear from time to time.
Then you should also consider that some noise may be observed and thus be present in your spectra.
Considering this, it's not totally obvious to assume that the "order" of the peaks will or won't be preserved between two consecutive scans.

But if we consider your example of two nearly isobaric peaks, you can indeed make the assumption that the appearance of a new peak between these two peaks is lowly probable.
What could happen more frequently however is that (depending on the instrument resolution and ion intensity), the two peaks may be merged into a single peak in some scans but not in other ones.
If the resolution is really good then you should be able able to distinguish your two peaks clearly and this pair should have a distance which is more or less conserved from one scan to another one.
But sometimes the data are full of surprises... so be careful with this kind of assumptions.

Best,

David