I have prepared a spectral library for a phosphopeptide enriched sample and I am generating my SWATH samples from similarly enriched samples. The problem is that when I use DIA-NN for the retention time alignment and quant, it doesn’t recognise the terminology of the spectral library annotated by ProteinPilot. DIA-NN recognises Unimod:21 for phosphorylation, but PP uses phospho(Tyr) etc. Other than changing the data dictionary to get around the mismatch, anyone have any suggestions for how I might resolve this? Thank you, Roz
Author
Roz Jenkins
Member ResolvedTags
May 25, 2022
ProteinPilot phosphopeptide library and DIA-NN
4 Comments
Submit a Comment
You must be logged in to post a comment.
Thanks for the question Roz, I have run into this as well. I always prefer to do a Thorough search with ProteinPilot software to get the broadest search space but then I did run into your exact issue.
• One workflow option is to process the data with Thorough in ProteinPilot software – then import into SWATH 2.0 micro app to convert your group file into an ion library. At this point of import, you can select Exclude Modifications which will strip out all but the common modifications. This library will then work with DIA-NN. But that of course will strip out the phosphorylations so doesn’t help you.
• When you use SWATH 2.0 micro app to convert the group file to an ion library, it does translate mods into the three-letter codes (TLCs) and DIA-NN does support some of the common TLCs. But we are not sure which ones work and which ones don’t so it would be a bit trial and error to figure out and manually edit in the ion library.
• There is a command in DIA-NN called –full-unimod, you can read up about it in the command line tool section of the online manual for DIA-NN. Looks like this disables the conversion of the modifications to the UniMod format. If all you are doing is using DIA-NN to extract quant of your peptides maybe this will work (just don’t have DIA-NN try to also build a library, as for that it needs to recognize all the mods). Because the proper fragment ions are already defined in your upfront ion library, it kind of doesn’t matter if the mod is recognized right? Note we haven’t tried this yet, but would love to hear back if you have success with this strategy!
• Note there is also a command called –mod [name],[mass],[optional: ‘label’] which lets you declare modification, so you could use this to add the key mods from your ProteinPilot software search into your DIA-NN processing.
Perhaps someone else with chime in with ideas!!
Christie
Hi Christie,
Many thanks for getting back to me. DIA-NN didn’t recognise ‘Pho’for phosphorylated, so no joy there. I tried converting all the PP three letter codes into the Unimod codes, but DIA-NN didn’t recognise those, even UniMod:21 which it was already using for phosphorylation! I’m also going to try to remove all reference to modifications in the ion library in the hope that DIA-NN will then detect the phospho groups itself during the alignment and library build. I will have a look at the command lines you mentioned as well. Thank you, Roz
Hi Christie,
I have had some success with this issue. I checked off the modifications on the DIA-NN dashboard. I then included Pho along with 6 other common modifications (carbamidomethyl, oxidation etc.) as –var-mod: DIA-NN recognised all the mods except Pho and converted them to their UniMod codes. However, it still listed the quants for the phosphorylated peptides, just labelled Pho instead of UniMod:21. This is just fine and should enable me to go ahead with my analyses. For any other users, the syntax is –var-mod[Pho],79.966,STY,label.
Best wishes,
Roz
PS: this requires that you perform a thorough search in PP but with biological modifiactions turned off and phosphorylation emphasis checked on. Then strip out all the other less common modifications (eg Delta:H(2)C(2)) from your ion library – find/replace in Excel works fine, just a bit tedious.