MRM Newbie Question.

[mburleson]

Hello!

I am a total newbie when it comes to performing MRM on triple quad LC-MS systems. So far, I am entirely self-taught with setting up the methods in our software. I was hoping someone could answer some questions that I have?

1. Should there a single spike in the TIC for each transition? For example, if analyzing reserpine and monitoring four transitions, would there be four distinguishable spikes in the TIC? I'm asking this for both clarity and for attempting quantitative work.

2. Our software (TSQ Quantum Ultra with Xcalibur 4.3 and Tune Master) has information that seems conflicting to me. When I do MS/MS optimation within Tune Master, it optimizes for MRM at a scan time of 0.1 s. While reading the help file in Xcalibur's Instrument Method tab it says the default scan time is 1.00 s which seems like overkill. Is the data below that is labeled as real data acquired with too fast of a scan time compared to that I found within a help file?



Thanks to anyone who can help!

[JMB]

I wrote a detailed response, but apparently did not post it.

Will repeat later today.

Regards,

JMB

[mburleson]

I wrote a detailed response, but apparently did not post it.

Will repeat later today.

Regards,

JMB

I look forward to your response. I appreciate your time!!

[JMB]

Q. 1) spike is a "bad" word in MS circles. It implies a noise event/electrical discharge etc. Here you are expecting 4 responses, if you set up 4 MRM expts in the datafile. They should occur chromatographically aligned beneath each other.

Q. 2) I cannot tell you exactly what the scan acquisition details in the HELP files were. However, I can point out some details on their printouts that you should be aware of,
a) SM 5G means that a 5-point gaussian smooth has been applied to the mass chromatogram. This done to make the peak look "GOOD" with a continuous peak shape. As a matter of interest, try applying in succession a 1-point, 3-point, 5-point Gaussian smooth to your own data file; note how the very spiky chromatogram becomes smoother and smoother, and how the area of the peak (peak counts) changes. Try the effect of other smoothing algorithms and compare.
b) The analyst is looking at the loss of 45 amu from the ion at m/z 465.3.
465.3 - 45 = 420.3; a peak centroided at m/z 420.3 would be 419.8 to 420.8 (1 amu wide at unit mass resolution).
However, the analyst has set his detection range at m/z 419.3 to 421.3 (this is 2-amu wide ) and gives a greater response than the 1-amu window above.
c) The same wider than 1-amu detection window is used for the D4 internal std. shown in the lower HELP file.
Similarly, the analyst can widen the width of the parent ion that undergoes MS/MS fragmentation; instead of m/z 464.8 to 465.8 (as above) for the 12C, 1H,14N etc molecule [M+H] , the 2H, 15N, 15N isotopic molecule at [M+H+1] could be included to increase the response.

Data smoothing and using expanded width detection windows are widely used, esp. in low-level quantitation by MS.
Note also that the deutero- analog mass chromatogram maximises just 1-2 scans ahead of the compound being quantified---this is a general effect.

Looking at the mass chromatogram profiles in the HELP files it looks like most of the signal lies in the time range 4.8 to 5.0 min, i.e. 12 secs wide. you need a certain minimum number of scans to define the chromatographic peak shape.
1-secs scan times would give you only about 3 scans at each of 4 x MRM expts. NOT ENOUGH
0.5 sec.......about 6 scans NOT ENOUGH
0.25 sec.....about 12 scans JUST enough ?/!
0.1 sec..... about 30 scans EXCELLENT to define peak shape.

Suggest that to get a feel for this that you run 3 expts successively with your 4 x MRM at each of the time settings.

Find the best time setting for data acquisition, then run 3 expts successively with the parent ion set at m/z 609.5 (1-amu wide), then 2-amu wide then 3-amu wide. See how the intensity of m/z 174 changes.
NB don't forget to expand m/z 174 if you widen the mass width on the parent ion.

d) Looking at your datafile it looks like you have set an extremely low width (174.001 - 173.999 = 0.002 amu). If the daughter ion actually is at m/z 174.0, you should set m/z 173.5 to 174.5 scan range. if at m/z 174.2, then m/z 173.7 to 174.7
NB There will be cases when, because of interfering compounds in the sample solution you may have to use very narrow mass windows.

We look forward to seeing your progress.
James Ball has a LOT of expertise in this area and if he posts, you should listen!

Regards,

JMB

[mburleson]

Q. 1) spike is a "bad" word in MS circles. It implies a noise event/electrical discharge etc. Here you are expecting 4 responses, if you set up 4 MRM expts in the datafile. They should occur chromatographically aligned beneath each other.

Q. 2) I cannot tell you exactly what the scan acquisition details in the HELP files were. However, I can point out some details on their printouts that you should be aware of,
a) SM 5G means that a 5-point gaussian smooth has been applied to the mass chromatogram. This done to make the peak look "GOOD" with a continuous peak shape. As a matter of interest, try applying in succession a 1-point, 3-point, 5-point Gaussian smooth to your own data file; note how the very spiky chromatogram becomes smoother and smoother, and how the area of the peak (peak counts) changes. Try the effect of other smoothing algorithms and compare.
b) The analyst is looking at the loss of 45 amu from the ion at m/z 465.3.
465.3 - 45 = 420.3; a peak centroided at m/z 420.3 would be 419.8 to 420.8 (1 amu wide at unit mass resolution).
However, the analyst has set his detection range at m/z 419.3 to 421.3 (this is 2-amu wide ) and gives a greater response than the 1-amu window above.
c) The same wider than 1-amu detection window is used for the D4 internal std. shown in the lower HELP file.
Similarly, the analyst can widen the width of the parent ion that undergoes MS/MS fragmentation; instead of m/z 464.8 to 465.8 (as above) for the 12C, 1H,14N etc molecule [M+H] , the 2H, 15N, 15N isotopic molecule at [M+H+1] could be included to increase the response.

Data smoothing and using expanded width detection windows are widely used, esp. in low-level quantitation by MS.
Note also that the deutero- analog mass chromatogram maximises just 1-2 scans ahead of the compound being quantified---this is a general effect.

Looking at the mass chromatogram profiles in the HELP files it looks like most of the signal lies in the time range 4.8 to 5.0 min, i.e. 12 secs wide. you need a certain minimum number of scans to define the chromatographic peak shape.
1-secs scan times would give you only about 3 scans at each of 4 x MRM expts. NOT ENOUGH
0.5 sec.......about 6 scans NOT ENOUGH
0.25 sec.....about 12 scans JUST enough ?/!
0.1 sec..... about 30 scans EXCELLENT to define peak shape.

Suggest that to get a feel for this that you run 3 expts successively with your 4 x MRM at each of the time settings.

Find the best time setting for data acquisition, then run 3 expts successively with the parent ion set at m/z 609.5 (1-amu wide), then 2-amu wide then 3-amu wide. See how the intensity of m/z 174 changes.
NB don't forget to expand m/z 174 if you widen the mass width on the parent ion.

d) Looking at your datafile it looks like you have set an extremely low width (174.001 - 173.999 = 0.002 amu). If the daughter ion actually is at m/z 174.0, you should set m/z 173.5 to 174.5 scan range. if at m/z 174.2, then m/z 173.7 to 174.7
NB There will be cases when, because of interfering compounds in the sample solution you may have to use very narrow mass windows.

We look forward to seeing your progress.
James Ball has a LOT of expertise in this area and if he posts, you should listen!

Regards,

JMB

THANK YOU SOOOOO MUCH for this!!

I'll be sure to avoid the term "spike." It was a bad word when doing Raman in grad school too, haha.

Thank you also for the breakdown of the scan times and number of scans. I've seen anywhere between 12 - 20 scans/peak is recommended. Thermo said 14 when I gave them a call regarding this. I will definitely give the experiments a try at the different scan times to get a feel for it!

The 0.002 amu width seemed crazy small to me as well. That was the default value in the HELP file. Your suggestion of 0.5 amu makes entirely more sense to me than such a narrow window - at least for now while trying to get a feel for this.

[JMB]

Everything you wanted to know about MS, but were afraid to ask......

https://www.rsc.org/images/MS2new_tcm18-102519.pdf

.....eine kleine nacht-MuSik for the coming winter months.....

especially p. 7 et seq.

Regards,

JMB

[mburleson]

Everything you wanted to know about MS, but were afraid to ask......

https://www.rsc.org/images/MS2new_tcm18-102519.pdf

.....eine kleine nacht-MuSik for the coming winter months.....

especially p. 7 et seq.

Regards,

JMB

This was AMAZING!! Thank you!! You have been such a tremendous help and I'm beyond appreciative.

[James_Ball]

I am kinda late to the conversation but I do agree with JMB, especially on the narrow mass window. Unless you are running a high resolution instrument like an Orbitrap and you have a good concentration to look for I would avoid anything that narrow. You would need extremely good stability to run that successfully.

For something like pesticides in drinking water I can get away with even up to 1 amu wide for parent and daughter ions, for a waste water or soil it may be good to narrow that up but you are also probably not needing as low of a detection limit.

Also I tend to run short dwell times. 10msec are normally enough for each MRM pair and you can run as many pairs as you can that keeps you in that 10-20 scans per peak range. 10 pairs at 10ms gives you 0.1 seconds per scan, and unless your peaks are less than 1 second wide you should still be good. Most modern instruments can handle even 5msec dwells without loss of sensitivity or stability. Unless you are doing changes in polarity, then you have a little more overhead in the scans.

If you start with wide amu windows and see little interference you are good and will have better stability and sensitivity. If you begin to have interfering peaks then try narrowing the windows until you begin to lose sensitivity or stability in the measurements. Part of good method development is making a good method then seeing what changes it takes to break it, then optimize within those bounds for the best reproducibility possible.

[mburleson]

I am kinda late to the conversation but I do agree with JMB, especially on the narrow mass window. Unless you are running a high resolution instrument like an Orbitrap and you have a good concentration to look for I would avoid anything that narrow. You would need extremely good stability to run that successfully.

For something like pesticides in drinking water I can get away with even up to 1 amu wide for parent and daughter ions, for a waste water or soil it may be good to narrow that up but you are also probably not needing as low of a detection limit.

Also I tend to run short dwell times. 10msec are normally enough for each MRM pair and you can run as many pairs as you can that keeps you in that 10-20 scans per peak range. 10 pairs at 10ms gives you 0.1 seconds per scan, and unless your peaks are less than 1 second wide you should still be good. Most modern instruments can handle even 5msec dwells without loss of sensitivity or stability. Unless you are doing changes in polarity, then you have a little more overhead in the scans.

If you start with wide amu windows and see little interference you are good and will have better stability and sensitivity. If you begin to have interfering peaks then try narrowing the windows until you begin to lose sensitivity or stability in the measurements. Part of good method development is making a good method then seeing what changes it takes to break it, then optimize within those bounds for the best reproducibility possible.

Thank you for your insight! I played around with adjusting the peak widths and scan times to assess the spectral quality. This forum has proven to be a lifesaver!

[James_Ball]

I am kinda late to the conversation but I do agree with JMB, especially on the narrow mass window. Unless you are running a high resolution instrument like an Orbitrap and you have a good concentration to look for I would avoid anything that narrow. You would need extremely good stability to run that successfully.

For something like pesticides in drinking water I can get away with even up to 1 amu wide for parent and daughter ions, for a waste water or soil it may be good to narrow that up but you are also probably not needing as low of a detection limit.

Also I tend to run short dwell times. 10msec are normally enough for each MRM pair and you can run as many pairs as you can that keeps you in that 10-20 scans per peak range. 10 pairs at 10ms gives you 0.1 seconds per scan, and unless your peaks are less than 1 second wide you should still be good. Most modern instruments can handle even 5msec dwells without loss of sensitivity or stability. Unless you are doing changes in polarity, then you have a little more overhead in the scans.

If you start with wide amu windows and see little interference you are good and will have better stability and sensitivity. If you begin to have interfering peaks then try narrowing the windows until you begin to lose sensitivity or stability in the measurements. Part of good method development is making a good method then seeing what changes it takes to break it, then optimize within those bounds for the best reproducibility possible.

Thank you for your insight! I played around with adjusting the peak widths and scan times to assess the spectral quality. This forum has proven to be a lifesaver!

Very true, I have learned a lot from this forum. I have been self taught on most of my analytical work and LCMSMS was probably the most difficult. There is a wealth of information here. If you learn something new, pass it on, we all benefit from it

[lmh]

also late to the party, and warning: I have not used Xcalibur with triple quads, but my understanding of Xcalibur is that unfiltered TIC will always be an unfiltered plot of everything the mass spec is doing. If you've got 4 transitions from reserpine, the TIC will show each in turn, which means the signal will zig-zag up and down, as each transition will give a different signal-per-concentration. Don't get rid of this by splatting a smoothing function on top of it! The right thing to do is to pick a transition that you want to use, and set the filter in the quantitative method.

[mburleson]

also late to the party, and warning: I have not used Xcalibur with triple quads, but my understanding of Xcalibur is that unfiltered TIC will always be an unfiltered plot of everything the mass spec is doing. If you've got 4 transitions from reserpine, the TIC will show each in turn, which means the signal will zig-zag up and down, as each transition will give a different signal-per-concentration. Don't get rid of this by splatting a smoothing function on top of it! The right thing to do is to pick a transition that you want to use, and set the filter in the quantitative method.

Thank you for your comment and insight! I will be sure to do that!