Formal Representation of Post Translational Modifications

[huckste...@gmail.com]

Hello,

Firstly, I am having problems with the BioPax download of the entirety of Homo Sapiens. Does anyone know when that will be fixed?

Secondly, is there any formal representation of post-transnational modifications that I could access? I cannot seem to find a single official representation/ way to extract a ptm from a pathway represented in gpml.

Any guidance would be incredibly helpful!

Thanks!

[Alex Pico]

I can confirm that BioPax download is broken, but I don’t know that status on its fix…

I can speak to the PTM information, however. This is still experimental and we are keen to know how you would use it. This would help us decide on the best approach to encode it and make it accessible. Our current use cases led us to encode information like this: "parent=G8JLH9; position=Y705; ptm=p; direction=u” for each PTM, inspired by PhosphoSitePlus. At this point, the information is in the GPML for a handful of pathways (that we did by hand). Once we settle on a format, then we would automate it, but we are not there yet.

So, what PTM information would you LIKE to see? How would you LIKE to access it?

 - Alex

--
You received this message because you are subscribed to the Google Groups "wikipathways-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wikipathways-dis...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/wikipathways-discuss/d6720c5e-cde8-4516-9758-11e2e0fb0724%40googlegroups.com.

[hannah huckstep]

Hey Alex, 

I do not work with the GPML format so I cannot really speak to it. 

My plan was to parse each file manually by searching for a key term such as “phospho” and get the uniprot ID associated. 

The main information I am looking to extract is the Uniprot ID, the position of the modified residue, the actual residue (eg. S,T or Y) and type of modification. 

Of course any additional information would be great, such as the protein cellular location, isoform, and direction. 

Something to consider is the nomenclature of your modification type. Currently if you look at the BioPax for Reactome and PhosphoSitePlus they differ in what names they call some modifications but are consistent in phosphorylation nomenclature. It would be ideal if you followed suit so that extracting modifications can use a generic script. Also both databases separate out phosphorylation into S, T and Y but Reactome has an additional category for phosphorylated proteins where the modified residue position is unknown. 

Reactome Phosphorylation types: 

phosphorylated residue

O-phospho-L-threonine

O4'-phospho-L-tyrosine

O-phospho-L-serine

PhosphoSitePlus Phosphorylaiton types: 

O-phospho-L-serine

O-phospho-L-threonine

O4'-phospho-L-tyrosine

Generally helpful things are when the ptm is attached to the isoform's uniprotID not the generic ID. Also with the format you have proposed I would have to use a regex to separate the amino acid from the location but having seperate variables for the position and modified residue would be helpful. 

Thanks for your help! 

Hannah 

You received this message because you are subscribed to a topic in the Google Groups "wikipathways-discuss" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/wikipathways-discuss/1skI5K9nWzA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to wikipathways-dis...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/wikipathways-discuss/DE9AB7B0-1A22-4B4C-8743-8BC5CECD187B%40gladstone.ucsf.edu.

[Alex Pico]

Thanks for the feedback and suggestions!

If you search each file for "ptm=p“ then you should find all the annotated phosphorylation events on our pathways so far. You’ll want to grab the other fields contained by the same “Comment” tag to get parent and position, etc.  In fact, some of the comment have more details like surrounding AA sequence, phosphosite group ID, and parent protein symbol.  

Again, there aren’t a lot since we’re doing it manually still. We might have the list of the pathways with these annotations. I’ll ask Kristina to send that next...

 - Alex

To view this discussion on the web visit https://groups.google.com/d/msgid/wikipathways-discuss/D263B1AD-3662-4736-8858-742598115A72%40gmail.com.

[Kristina Hanspers]

Hi Hannah,

Attached is a list of WPIDs that have annotated phosphorylation sites. These are all human pathways except for the very last one, which is mouse.

Regards,

Kristina