Output Table Issues - Intensity vs. Unique Intensity vs. Total Intensity

[mamccabe]

Hello,

I have been using MSFragger for processing large proteomics datasets but am unclear on how to interpret some of the results. I have referenced the output wiki page (https://github.com/Nesvilab/philosopher/wiki/Combined-protein-reports) but I believe that, at least for the combined protein reports, the stated descriptions do not match what the columns for " Intensity", " Unique Intensity", and " Total Intensity" actually contain.

The first issue is that on the wiki page, the columns are listed as " Total Intensity", " Unique Intensity", and " Razor Intensity". Using the most up-to-date versions of MSFragger (3.3) and Philosopher (4.0.0), my outputs contain a " Intensity" column rather than a " Razor Intensity" column. This could be due to settings but I have not found documentation on how to change this. It could be reasonably assumed that the " Intensity" column represents the razor intensities, however I will describe below why I am not sure this is the case.

The second issue is that the column values apparently do not match their descriptions. In the second sheet of the attached Excel file, you can see 3 columns where the unique, total, and standard intensity values are compared for a single sample. Based on the descriptions in the wiki, you would expect that Total Intensity (all matching peptides, regardless of uniqueness) would always be the same or higher than the unique intensity (only peptides unique to that protein). However, in the third line down (as well as in many other cases), you can see that the unique intensity is higher than the total for that protein. Additionally, one would assume that razor intensity (peptides assigned to a protein using the razor rule) should always be lower or the same compared to both the unique and total intensities, but there are cases where the razor intensity is higher than both of these values.

In summary, there are no consistent relationships between the three columns for intensity (and by extension spectral matches) which logically match the descriptions provided on the wiki. I'm unsure if this is a description issue or an issue with the actual outputs, but I'm sure clarification would be useful for me and many others.

Thank you!

Max

Philosopher Support Upload.xlsx

[sarah-haynes]

Hi Max, we need to update that page, thanks for noticing. For now, can you look at the combine_protein.tsv description? This one has been updated.

Sarah

[mamccabe]

Great, that does significantly clarify what the columns should contain! Thanks for pointing me in the right direction.

However, I still am a little unsure about the second portion of the issue, where the unique intensity is sometimes greater than the total intensity in the output file. It seems like this should never be true given the updated descriptions as well. I checked if the same logical issue is the case for spectral counts, and everything seems to be correct there (i.e. both other spectral count columns are always lower than the total column). However, there seems to be an issue still with how the unique and total intensities are calculated based on their relationship to one another.

I can transfer this to the issues section if it would be better fitting as a bug now!

Thanks,

Max

[prvst]

@mamccabe We did some changes in the way the reports are put together, specially when you have a combined experiment. We will release a new version soon.