Unable to train peak scoring models

Unable to train peak scoring models mmidha  2019-01-10

Hi Brendan,

I am processing the DIA experiment on Skyline I am unable to reintegrate and train the peak scoring model based on mprophet after the processing is complete. It seems there are not many available features scores to do that, only first two features are enabled but rest are disabled (please see attachments). Kindly request you to look into the issue.

Thank you


Nick Shulman responded:  2019-01-10
Those disabled scores are things that cannot be calculated for some of the peptides in your document.

You should select the row in the grid for one of those disabled scores.
Then, click on the "Feature Scores" tab, and Skyline will show you a bar graph of the score values in your document. If the score is present on some but not all of your peptides, then there will be one bar in the graph labeled "unknown". If you click below that bar, there are some binoculars which tell Skyline to give you a list of all of the peptides which did not get that score.

There are many reasons a score might not be available. For instance, library intensity dot-product will be disabled if the peptide was not found in the library, or peptide has fewer than 3 MS2 transitions

I have actually never seen "Intensity" disabled. It might be obvious to you once why it's disabled once Skyline shows you the list of peptides for which it is disabled. (Maybe those peptides have no chromatograms).

If you can't figure it out, you should send us your Skyline document.
In Skyline, you can use the menu item:
File > Share > (complete)
to create a .zip file containing your Skyline document and supporting files including extracted chromatograms.

You can upload that .zip file here:
-- Nick
Brendan MacLean responded:  2019-01-10
Also, the Advanced Peak Picking Models tutorial includes instructions on how to look at which targets are causing scores to be disabled. I highly recommend learning that skill for working with mProphet models.