Conversion from Skyline to MSstats for PRM analysis

[Giovanni Canil]

Dear MSstats group,

I am encountering some difficulties in importing PRM data into MSstats. What I am doing is creating the Report 'Transition Result' from Skyline, which has the info on Peptide, Replicate, Precursor, Fragment Ion, Area etc, and importing an annotation file with the usual architecture, then submitting both of them to the 'SkylinetoMSstatsFormat' function, but I get the error:

'Error in .checkDDA(input): **Please check precursors information. If your experiment is DIA, please remove the precursors. If your experiment is DDA, please check the precursor information.'

Since I believe PRM is quantified at the fragment level, do I have to get rid of the precursor column? Or is there another Report from Skyline I need to export (which I could not find)?

Thank you in advance for your reply and for developing such a useful tool,

Giovanni

[Anthony Wu]

Hi,

I know MSstats supports conversions of datasets produced from DIA, DDA, and SRM acquisition types, but I'm not too familiar with PRM.  As a first step, could you provide us the sample data and the annotation file so we could reproduce the error?  

Thanks

Tony

[Anthony Wu]

Hi Giovanni,

Thanks for sharing the data to my work email.  I'll take a look later this/next week.  I saw you added an additional question which I'll add to this google group thread:

As a side question, I wanted to ask you what is the process MSstats does to output the protein intensity with DIA data. If I understood correctly, the process is based on the summarization of product ions intensity, without dealing with precursor ions, am I right?

Yes, your intuition is generally correct.  

To be more specific on what MSstats does (ref: section 2.2 in page 10 of the MSstats manual) - once raw data (e.g. data from Skyline) is converted to MSstats format, the combination of values in the columns peptide sequence, precursor charge, fragment ion, and product charge uniquely determines a feature.  For each protein, MSstats determines the protein's intensity by taking a summarization of the intensities of the protein's features.

Tony

[Anthony Wu]

Hi Giovanni,

I took a look at your dataset, and I have a few questions:

1. I saw for the "Fragment Ion" column, there are rows with the values of "precursor", "precursor [M+1]" and "precursor [M+2]".  Could those rows be removed?  From my understanding of SRM/PRM, the first MS scan deals with precursor ions, and the second MS scan handles the fragment ions, but we only care about intensity measurements from fragment ion scans (ref).
2. Are you able to find an "MSstats format" report in Skyline which you can export your data?  From what I know, there should be an MSstats format report in Skyline

Thanks,

Tony

[Anthony Wu]

Hi Giovanni,

Pasting your previous email message here for visibility for other MSstats team members:

Thank you for looking at the data. To answer your questions, I was able to export the Report in a format that MSstats could read, I imported that and began the processing. However, when I get to the point of the Comparison, the ComparisonResult table gave me NA values for the columns SE, Tvalue, pvalue, adj.pvalue and 0 for DF. The issue column is all NA and the Missing and Imputation Percentage are 0. 

I took an initial look and this issue seems to originate from your dataset only having 1 bioreplicate per condition.  In order for the comparisonResult table to have meaningful information, you'll need to have at least 2 bioreplicates per condition.  

Best,

Tony

[Anthony Wu]

Hi,

May I ask you another question on the transformation from Proteome discoverer to MSstats?

When I import the data with the 'PDtoMSstatsFormat' function, I get the warning 'Warning message:

In max(common$n_common) : no non-missing arguments to max; returning -Inf'and the quant file has NA under the Condition and Bioreplicate columns. Can you comment on the two issues?I attached the annotation file and a light version of the csv PD report file.

In response to this question, below are the root causes of your problems:

1. NA for condition & bioreplicate - In your annotation file, all runs are formatted as "<run-id>_DIA", but in the PD report file, all runs are formatted as "<run-id>.raw". The values in the "Run" column of the annotation file need to match the values in the "Spectrum File" column in the PD report file.
2. Warnings - In your PD report file, there are blank values in the column "Precursor Abundance", causing the warnings to appear.

Best,

Tony