Separate files vs. merged files (.mgf) analysis

[Juan Camilo Rojas Echeverri]

Hi again,

I performed an export of various LC-MS/MS runs with Progenesis to get .mgf files for processing with SearchGUI+PeptideShaker. Progenesis pools all the MS/MS into a single .mgf file. However, for my downstream analysis I need the MS/MS of each file separately. I asked a colleague to write a short python script to split merged .mgf file into separate files based on the title of the spectrum. All the MS/MS spectra are accounted when the split files are compared so I don't think the splitting is the problem.

Nevertheless, I wanted to check if there would be a difference in the search. I was expecting no difference since is the same MS/MS spectra. But I'm getting different numbers of peptides and protein coverage.

Is there anything in the validation steps of PeptideShaker that you could see could cause reaching different results when the same data set is processed as one pooled file vs. separate files?

Thanks for the time and help.

JC

[Harald Barsnes]

Hi JC,

> Is there anything in the validation steps of PeptideShaker that you could see could cause reaching different results when the same data set is processed as one pooled file vs. separate files?

There reason for the results (potentially) being different is that the search results will be slightly different when looking at the whole data set at once compared to splitting it into smaller subset, given that this can affect the scores of the individual PSMs and the associated e-values. Which will again be picked up by PeptideShaker when processing the data.

Best regards,

Harald