Combining own DDA library with SWATH atlas library for DIA data moe  2016-09-04
Hi all,

I've read the Skyline iRT Retention Time Prediction tutorial but I wondered if I could get some help. I would like to combine an in-house DDA library (generated from a DDA experiment of depleted human plasma) with an external DDA library (e.g. Rosenberger library).

1) I've created my own human plasma DDA library using my own DDA experiment loaded into skyline via File - Import - Peptide Search

2) Additionally I've downloaded a custom library from SWATH Atlas (phl004_canonical_sall_osw.csv) from here, selected the proteins of interest (all proteins identified by Farrah et al in plasma) and generated a library in skyline via File - Import - Transition List

I now have the two libraries in skyline and I would like to align the retention times of both libraries to generate a more comprehensive library and get the most out of my DIA data.

If possible I aim to use ApoA1 to calibrate retention times of my own plasma DDA library and then use the calibrated retention times of ApoA1 and convert all retention times of the online SWATH atlas library.

Any help on how I can achieve this would be appreciated.

Brendan MacLean responded:  2016-09-05
Hi Moe,
Well, obviously you can combine the spectral library aspects of your two libraries easily enough in Skyline by checking both libraries in the Peptide Settings > Library tab. However, Skyline allows only a single iRT retention time library/calculator. So, you need to build that up, as described in the iRT tutorial, starting with at least some data which contains the standards you expect to use for your iRT regressions. Once you have done this, and you have enough peptides in your iRT library/calculator that new data sets have enough overlap with what is already in the library, you should be able to add from sources which do not necessarily contain your standards.

Just keep clicking the "Add" button in the Edit iRT Calculator form and choosing new sources of RT measurements. As long as they have enough information for Skyline to be able to make the conversion, this should work. These sources include 1) the current Skyline file with imported results, 2) spectral libraries, 3) other iRT calculators.

For #2, this will only work if the spectral library contains enough information, specifically retention times and source file assignments (need both to make valid regressions).

Luckily, #3 is probably your preferred route. You should be able to import the CSV files from the SWATH Atlas using File > Import > Transition list (we may soon also add File > Import > Assay Library for better name recognition). This operation should ask you to create both a spectral library and an iRT calculator from the data in the CSV file. If you get through that successfully, and the result is a new iRT calculator with enough overlap of another iRT calculator, then you should be able to combine the two using #3. I have done this myself in collaborative projects with the Aebersold lab. I know it works.

Hopefully that is enough to point you in the right direction. Sorry we don't have step-by-step instructions for this sort of thing yet. The iRT tutorial seemed long enough without covering absolutely everything that can be done.

Thanks for posting this question to the support board.