Issue with Unimod modification IDs needed for ProteomeXchange submission j hamey  2019-12-10
 

Hi there,

I'm having an issue submitting data to Panorama public that I would like to get a ProteomeXchange ID for.

In order to analyse my data in Skyline, I had to create custom modifications. In particular, I created two modifications that are essentially the same as two existing unimod modifications (accessions 1291 and 298), but which were not findable in Skyline.

The submission process says that the modifications must have a Unimod ID in order to get a ProteomeXchange ID, and that the remedy is to edit the modifications in Skyline and reupload everything. However, this seems like it would be untenable given the number of Skyline documents and peptides in my dataset, and the modification names in Unimod would also make the dataset more confusing. Is it possible to simply associate the data with the two relevant Unimod modifications rather than change every instance of these modifications?

Thank you for your help.

Josh

 
 
Brendan MacLean responded:  2019-12-10

Skyline has these modifications built-in:

298
"Methyl:2H(3) (C-term)"
"Methyl:2H(3) (D)"
"Methyl:2H(3) (E)"

1291
"Dimethyl:2H(6) (K)"
"Dimethyl:2H(6) (N-term)"

Are your needs somehow not covered by these, or were you unable to find them? When you define a modification the name field has a dropdown list. I assume you have found at least that much, or you would likely be complaining that Skyline has no concept of Unimod modifications. To find these more specialized modifications, you need to first choose the <Show all...> element which will then become <Show common...> to allow you to switch back to the shorter list.

If you then drop the list and start typing "Dimethy" Skyline will bring those modifications into view, and you will see the ones I have listed.

I hope this helps in the future. Vagisha may also be able to help with translation on the server-side, as she has done in the past for more common modifications that were defined in the past without assigned Unimod IDs.

--Brendan

 
j hamey responded:  2019-12-10

Thanks for the quick response Brendan.

You're right that these modifications are available as isotope modifications, however they're not present in the structural modifications drop-down list, which is what I have been using. Sorry I should have clarified though that I specifically need these modifications on Arginine, which is not found in either list, but is in Unimod (specificity definitions 5 and 3 for 298 and 1291). I tried to edit the specificity for one of these, however it then says that the modification doesn't match the Unimod specifications, and asks if I want to use non-standard settings (see attached screenshot). If I do this, will it still cause the same issue downstream when submitting for a PX ID?

And if this does satisfy requirements for getting a PX ID, is there any relatively straightforward way to convert every instance of the two modifications to the Unimod versions?

 
vsharma responded:  2019-12-10

Hi Josh,

Can you please send me the link to the folder that you are trying to submit to Panorama Public? You can email the link to me at vsharma@uw.edu if you don't want to post it here. I should be able to submit this to ProteomeXchange with the correct Unimod IDs but would like to take a look at the data first.

Was there a reason to add these as structural modifications in Skyline instead of isotopic modifications? They are classified as isotopic labels in Unimod.

Thanks,
Vagisha

 
Brendan MacLean responded:  2019-12-10

Strange. There is a Dymethyl(R) in the structural modifications along with K, N and N-Term. It seems a bit strange that 2H(6) is showing up in the isotope labeling modifications when 2H(4) shows up in the structural modifications. But, I guess that is because there are not 6 hydrogen atoms in the modification. Dymethyl:2H(4) shows up in the structural modifications as H'4C2. So, Dymethyl:2H(6) would need to be H'6C2 - H2. But, instead, it gets defined as H'6 - H6 in a way that needs to be added as an isotope modification on top of a normal Dymethyl modification defined as H4C2.

Yeah. It is a bit confusing, but it all gets worked out by a program in our project that parses the Unimod XML and comes up with the structural and isotope modifications, and, yes, it has been a couple years since we did this. But, this is the first serious issue to come of that.

Thanks for pointing this out. Vagisha will need to work with you on the PX ID.

--Brendan

 
j hamey responded:  2019-12-10

Thank you both!

Vagisha, I will send you an email with the link.