Secondary ID mapping service in Bridge DB

[DeniseSl22]

Moving this issue from pathvisio github to BridgeDB.

DeniseSl22 commented on Jul 31 Hi all, (again about the library, sorry about that ;)....)

In the Pathvisio library, secondary identifiers of ChEBI are included as well (Marvin found this for bicarbonate). I think for a new Pathvisio release, we should remove these secondary IDs as a option; and replace all secondary ChEBIs with primary in WPs. Since the secondary IDs are not mapped via bridgeDB.

@marvinm2 @egonw @mkutmon @DeniseSl22

DeniseSl22 commented on Jul 31 After consulting with Jonathan, he told me that the library(search function in Pathvisio) uses the content of BridgeDB databases. So, if any secondary metabolites IDs are being found in the "library", it is because they are in the metabolites bridgeDB file. We do not need a new Pv release for this change.

Is this something we want to include in the bridge file (I have other examples, where I could not create a mapping, because the secondary ID was used in a dataset....)? Or do we just want the primary ones included, and another way to map the secondary to the primary IDs? @mkutmon

mkutmon commented 3 days ago @DeniseSl22 did you move this issue to the bridgedb repo? then I close it here. @DeniseSl22

DeniseSl22 commented a minute ago Nope not yet; I will do it right away :) But it is a problem which we do not have the solution for right now (according to Chris and Egon)...

[egonw]

@DeniseSl22, please:

1. read http://dx.doi.org/10.1371/journal.pbio.2001414 and summarize what they say about secondary IDs
2. read the BridgeDb paper and make a proposal of how we should tag IDs in BridgeDb about whether the are primary or secondary

[AlasdairGray]

@egonw the notion of primary and secondary identifiers is included in the OPS IMS implementation. Did this make its way back into the BridgeDb codebase?

[ariutta]

datasources.txt specifies primary and secondary datasources. Here are the headers for that file.