Peak Integration Issues

support
Peak Integration Issues sa825  2021-09-19
 

Hi,
I am looking at some peptides over multiple replicates and I find that Skyline is not constant in its assignment of the peak range of each peak from replicate to replicate.

Please have a look at the images attached. For replicate 4 in the image attached, skyline selects from 25.8s to 26.27s but for replicate 5, it only selects from 25.8s to just over 26s. Is there a way to make Skyline more constant in the way it does this?

I could manually go around adjusting each peak but that would introduce a lot of bias in my work. I also plan to use Skyline for a method involving 116 peptides across 85 samples, so manually adjusting each peak would be a lot. Moreover, if you look at the "Peak Area- Replicate Comparison" graph, you could easily make the mistake in thinking that a peptide is not being properly detected in a particular replicate (e.g. replicate 5) as its peak area appears so small in comparisons to the other replicates but that's not actually the case, a smaller peak area is just being selected by Skyline.

Final question, is there a way to change the dwell time in Skyline when I export a method? I am exporting methods for a Xevo TQ_XS.

Thanks,
Shimon

 
 
Nick Shulman responded:  2021-09-19
I do not know the answer to your method export dwell time question, but hopefully someone else on the support board does.

When Skyline is detecting peaks in a chromatogram, the first thing that Skyline does is find the highest intensity along the entire chromatogram. Then, Skyline figures out the full-width-at-half-max from that maximal point. The next step of peak finding is to calculate the second derivate of each point along the chromatogram. However, since a chromatogram is a discrete set of points, and not a continuous function, the "second derivative" is actually a sort of weighted average of a rolling window of points along the chromatogram. The width of that rolling window is controlled by the width of that largest peak that Skyline found. This width calculation happens separately for each transition's chromatogram.

Another way to say this is that the width of the tallest peak on the chromatogram controls the amount of smoothing that happens when Skyline is looking for all of the other peaks along the chromatogram.

I cannot tell from your screenshot whether we are looking at the tallest peak along the chromatogram, or whether the tallest peak is on the part of the chromatogram which beyond the edges of the graph.

If you send us your Skyline document I could tell you more about why Skyline is doing what it is doing.
In Skyline you can use the menu item:
File > Share
to create a .zip file containing your Skyline document and supporting files including extracted chromatograms.

If that .zip file is less than 50MB you can attach it to this support request.
Otherwise, you can upload it here:
https://skyline.ms/files.url

Another thing that can give you a little more insight into why Skyline is doing what it is doing is to right-click on the chromatogram and choose "Transform > Second Derivative". The only important parts of the second derivative view are the places where the curve crosses the X-axis. Those are the inflection points of the original curve, and are the places where Skyline starts from when determining peak boundaries. But, the exact way that the second derivative was calculated depends on the width of the tallest peak along the chromatogram.

Sometimes it is possible to change the peak widths of that Skyline ends up looking for by changing the Retention Time Filtering at "Settings > Transition Settings > Full Scan" so that Skyline extracts full length chromatograms. In that way, the tallest peak along the chromatogram might end up being something different.

I am not sure whether any of this information will be helpful.
I think the answer that usually works better is something about improving spray stability or chromatography.
-- Nick
 
sa825 responded:  2021-09-20
Hi Nick,
Thank you for your rapid response. Hoping someone else can answer the question about dwell time please.

I think I have a better understanding now of how Skyline chooses its peak width. However, looking at Replicate 5, I wouldn't say its actually selecting the tallest/most intense peak?
Regarding the last suggestion you made, the skyline file had already been set to "Include all Matching Scans".
Attached is my Skyline file so you can look at it better.

Thanks,
Shimon
 
sa825 responded:  2021-09-21
Nick?
 
Nick Shulman responded:  2021-09-21
Skyline does not actually choose the tallest peak along the chromatogram.
Skyline figures out the full width at half max of the tallest peak is and then Skyline uses that width to decide how much smoothing to apply to the chromatogram when calculating the second derivative.
In replicate 4 in your screenshot, the tallest peak is .092 minutes wide (25.916 - 26.008). In your replicate 5 screenshot, the tallest peak is 0.025 minutes wide (26.049-26.074).
This difference in width causes Skyline to be more likely in replicate 5 to think that noise marks the edge of a peak.

If you want to learn about how Skyline uses the second derivative as the starting basis of peak detection, it's mentioned a little bit in Figure 2 in Lindsay Pino's "Skyline ecosystem" paper:
https://analyticalsciencejournals.onlinelibrary.wiley.com/doi/full/10.1002/mas.21540

I hope this makes sense.
-- Nick
 
sa825 responded:  2021-09-22
Hi Nick,
Thank you for your response and your explanation.
Is there any way to overcome the issue I'm facing i.e. to cause Skyline to maintain a specific peak range from replicate to replicate?
Thanks,
Shimon
 
sa825 responded:  2021-09-26
Nick?
 
sa825 responded:  2021-10-06
Nick?