Reintegration of analyte (light) affects ISTD integration (heavy)

support
Reintegration of analyte (light) affects ISTD integration (heavy) dawn.r.dufield  2017-08-18 13:28
 
Many times for a low level analyte (peptide or small molecule) quant, you have to adjust the integration parameters to get a good peak. Currently I only see a way to "manually" integrate a new peak, which although is not ideal as hard to reproduce etc is reasonable, however I have found that when I manually integrate the light or analyte peak, it also changes my heavy or ISTD peak which in most cases since its typically a larger more easily integrated peak we do not want to do that. Can you please comment on if there is a way around this. I will attach a screen shot of what I'm talking about.
 
 
Brendan MacLean responded:  2017-08-18 14:51
Hi Dawn,
We have moved more and more to the idea that the integration ranges should be the same for the ISTD and the analyte. Many people who started out making this same assertion have since decided they agree with Mike MacCoss that the two should be integrated the same and that generally, background subtraction takes care of the difference you describe. That being the case we have made this a very strong default making it very easy to achieve.

However, we did leave one case where they are not matched, and that may be enough for you on this small molecule experiment, and that is when you have single transition selected and you are in single transition viewing mode (Edit > Transitions > Single). In this mode any integration changes you make will only apply to that transition and not be forced on any others.

There is also a way to define a peptide isotope modification that has "unknown" retention time impact on the peptide, and reference precursor pairs where the heavy peptide has this type of modification are not forced to have identical RT boundaries, but I don't think this is used very much, and doesn't seem useful with our small molecule support.

On integration adjustment, you may also be missing that you can hover the mouse cursor over the dashed lines that denote the integration boundaries and the mouse cursor will become an "east-west" splitter cursor. If you click and drag when that happens you can adjust that single boundary, rather than "integrate a new peak", which you typically do by clicking and dragging the entire range beneath the y-axis. This second option is frequently the most convenient when both boundaries are not quite where you want them, but it is definitely possible to adjust either boundary from where they start by clicking and dragging on those boundary lines themselves.

Hope this helps. Thanks for your interest in Skyline.

--Brendan
 
dawn.r.dufield responded:  2017-08-18 15:23
Hi Brendan, thanks for the response, I cannot seem to get the first option to work. I currently am in single transition mode and when I hover and drag or reintegrate, both seem to change the heavy (ISTD too). Maybe I didn't set up the ISTD correct for this small molecule application. It is the same file I sent earlier. Maybe you can take a look.

I basically have been learning by clicking around. Do you have a good tutorial you can point me to where some of this might be discussed?

Thanks
Dawn
 
dawn.r.dufield responded:  2017-08-18 15:36
There will also be several cases where we use an analog ISTD and this again will not want to have the same RT or integration starts and stops.

Thanks
Dawn
 
Brendan MacLean responded:  2017-08-18 16:00
Hi Dawn,
That we do support with Surrogate Standards. Others have already requested it. (I will get the developer that implemented these to post instructions on them.)

So, I suppose, if you want to be able to integrate your analyte and standard differently, you could always put them in separate molecule elements and link the heavy as a surrogate standard. Obviously, a little less than ideal, but that would allow you to achieve what you are looking for.

I see that there is a bug in how single-transition integration is handled now. If you have 3 transitions in the same precursor, you can integrate each separately using the method I described. However, if that precursor is part of a group with other precursors, they will all adopt the integration boundaries of the most recent change. I will do my best to fix this for the next Skyline-daily. It pretty clearly makes no sense to have this happen.

Thanks for your feedback.

--Brendan
 
Nick Shulman responded:  2017-08-18 16:47
Here is the tip page that I wrote for how to use surrogate standards.
https://skyline.ms/wiki/home/software/Skyline/page.view?name=Surrogate%20Standards

Let me know if anything is not clear and I will update the page.
 
dawn.r.dufield responded:  2017-08-18 17:17
Thanks guys, I will give it a try. Appreciate the quick response.
 
dawn.r.dufield responded:  2017-08-18 20:42
Hi Nick,

I didn't get it to work quite right. I added a new small molecule and labeled it in the transition list as surrogate std. Then added the normalization column in the doc grid. In the doc grid I have a curve for the analyte with all the areas and then I have the same replicates for the surrogate STD. There does not appear to be a way to put a different (ie constant) analyte concentration for the surrogate STD. It keeps whatever I put on the analyte peak above. Also I get NaN for the conc and Accuracy. If I put the normalization method for the ISTD to normalize to surrogate (doesn't seem right) since it is the surrogate, then I get numbers for the analyte, but it does not change when I change the integration on the surrogate so clearly it is not normalizing. Not sure what else I did wrong. Any other suggestions?

Dawn
 
Nick Shulman responded:  2017-08-18 21:23
I am not sure I understand what you did.
Can you post your .sky.zip file?
You can either attach it to this support request (if the file is less than 50MB) or upload it here: https://skyline.ms/files.url
 
dawn.r.dufield responded:  2017-08-19 09:44
Ok I will attach here. Basically I added a new small molecule which I assigned as the surrogate STD. I then "normalized" to it in the doc grid, but when I change the integration of the ISTD (surrogate STD) the % accuracy does not change so I dont' think it is working as it should.

Let me know what you see.
 
Nick Shulman responded:  2017-08-20 16:13
Skyline is correctly normalizing by dividing by the surrogate standard peak area. However, it looks like there is a bug in that Skyline does not notice when the peak area of a surrogate standard has changed, so the normalized area never gets recalculated.

Until we fix this bug, the best way to work around this bug and trick Skyline into recomputing the normalized area is to change the "Standard Type" of the surrogate standard to "None", and then to change it back to "Surrogate Standard". (Saving, closing and reopening the Skyline document would work as well).

Sorry about this bug, we will try to fix it soon.

I think everything else is working correctly. In theory, you would not want to specify the analyte concentration for the surrogate standard, because the surrogate standard was spiked into each replicate at exactly the same concentration.

The reason that Skyline is giving you "#N/A" for the concentration of the surrogate standard (d9 ACH) is that the surrogate standard is marked as "Heavy", and on "Settings > Peptide Settings", the "Internal Standard Type" is also "Heavy". If you were to change that to "None", then Skyline would try to calculate concentrations for the surrogate standard, but it would definitely end up being the wrong number because the analyte concentrations on the Replicates are not correct for d9 ACH.