RT prediction not shown for all peptides/samples

support
RT prediction not shown for all peptides/samples ssaleh  2019-05-13
 

Hi Skyline team,

I recently run around 20 samples in quadruplicate on TSQ vantage (to monitor 12 peptides using MRM mode) and I am currently analyzing the data. When trying to upload the results into skyline I had the following error "unable to finish importing chromatograms because RT predictor linear regression failed". To solve this problem I had to switch off the RT predictor, this way the import was successful.
I then switched on the RT predictor , however the RT prediction was missing in a lot of peptides and samples in a random way (please check attached ppt ). Any idea what could be the reason?
Is there a way to share the sky.zip in a private mode?

Many thanks in advance!
Sara

 
 
Nick Shulman responded:  2019-05-13
If you have an iRT retention time predictor, then Skyline extracts chromatograms for your iRT standards. Then, Skyline attempts to do a linear regression between the retention time of the chosen peak for each standard peptide, and the iRT score for that peptide. If the R-squared for that linear regression is less than 0.99, then Skyline throws out one of the iRT standards and does the regression again. Depending on how many standard peptides you had to start with, Skyline is willing to throw out approximately 2 standards before Skyline says that the linear regression failed.

If you want to check what the failed linear regression looks like, you can use the graph:
View > Retention Times > Regression > Score to Run
On that linear regression graph, you want to make it so that you are only seeing the standards for the replicate that failed. So, use the following right-click menu items:
Replicates > Single
Calculator > {name of your iRT calculator}
Points > Standards

You will probably be able to see which peptide was not chosen at the correct time.

If Skyline is choosing the wrong peaks for some of your iRT standards, it might be that you have told Skyline the wrong set of transitions to look for.

I will send you email in case you want to privately send us your files.
-- Nick
 
Nick Shulman responded:  2019-05-13
Hi, Sara,

Thanks for sending me your data. I did not realize this was SRM data.
This is a bug in Skyline. If you have SRM data, then Skyline should not be refusing to import chromatograms like this just because the iRT linear regression failed. Chromatogram extraction should only fail if you have chosen "Use only scans within X minutes of predicted RT".
I am not sure when we will be able to fix this bug, but we will try to fix it soon.
-- Nick