search

PathVisio / pathvisio

PathVisio - pathway editor, visualization and analysis software
http://www.pathvisio.org
Apache License 2.0
22 stars 21 forks source link

What to do with (hybrid) pathways that contain genes of multiple species? #150

Open marvinm2 opened 4 years ago

marvinm2 commented 4 years ago

such as https://www.wikipathways.org/index.php/Pathway:WP1438

How can we annotate the species?

khanspers commented 4 years ago

Hi Marvin,

Since PathVisio only takes one species designation, we made this pathway human. If we had an influenza virus database we could have another copy of the pathway designated as influenza.

However, for data mapping you can use Cytoscape to map data to both human and viral data nodes.

Kristina

On Jan 30, 2020, at 6:45 AM, Marvin Martens notifications@github.com wrote:

such as https://www.wikipathways.org/index.php/Pathway:WP1438 https://www.wikipathways.org/index.php/Pathway:WP1438 How can we annotate the species?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/PathVisio/pathvisio/issues/150?email_source=notifications&email_token=AALADFIDBSS7AA7GUDTMD63RALRW3A5CNFSM4KNXGEGKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4IJ3M3LA, or unsubscribe https://github.com/notifications/unsubscribe-auth/AALADFMKAMOJ23AGG7ALWFLRALRW3ANCNFSM4KNXGEGA.

DeniseSl22 commented 4 years ago

Hi all; also discussed this issue with @mkutmon . Would it be an idea to extend the GPML model, so users can specify per datanode/interaction/lit. ref for which species this is relevant? Then the main species designation chosen in the beginning is implemented for all datanodes+interactions+lit.refs as initial values, but we give users an option to select a different species in the DataNode/Interaction properties panel. I could see some examples where we could use this information:

  1. Hybrid PWs (where info from multiple species is combined).
  2. Disease PWs (indicating some elements relationship to disease have only been established in certain animal model, for clarity. I think @fehrhart is adding this type of info now in comments).
  3. Make (meta)data for homology converted PWs more transparent; for example the LIPIDMAPS PWs have only been created for Mouse; I've added literature for mouse+human to complete the PWs, but no way now for automated retrieval which lit.ref belongs to which species.
  4. Create easier suited Systems Pharmacology models, see for example mevalonate PW with inhibitors, which has info for 2 species. I'm not stating this is the final model we would like to have (maybe easier to add the drugs later on through CyTargetLinker); but would be a nice extension for the kinetic semantic data model I've been working on (see poster).

We could even extend this approach later to specify in which organ/tissue/cell type a DataNode/interaction/lit. ref is relevant, which is something we wanted for quite some time if I remember correctly. This information would also allow us to align the GPML with Genome Scale Metabolic Modeling approaches, so we can use it for Fluxomics etc.