Align with UCSD Peter Rose's KG

[justaddcoffee]

We can discuss this on this ticket - Peter will align his data with the biolink model and KGX data format (nodes.tsv and edges.tsv) and then we can exchange data

@pwrose @deepakunni3 @wdduncan @cmungall

[wdduncan]

I also applied for access on https://hira-covid19.net/. Might also yield some interesting clinical information, and we can merge with Peter Rose's data.

[pwrose]

How do I find the biolink relationship types between nodes, e.g.

Genome -> Gene Gene -> Protein (I don't have any transcript information, so I need to link directly from gene to protein), e.g. Gene-ENCODES-Protein.

On Thu, Apr 9, 2020 at 2:01 PM Justin Reese notifications@github.com wrote:

We can discuss this on this ticket - Peter will align his data with the biolink model and KGX data format, and we can exchange data @pwrose https://github.com/pwrose @deepakunni3 https://github.com/deepakunni3 @cmungall https://github.com/cmungall

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Knowledge-Graph-Hub/kg-covid-19/issues/78, or unsubscribe https://github.com/notifications/unsubscribe-auth/AA53AEB4TMQNKVCJWSFJJQ3RLYZR5ANCNFSM4MFAW4XA .

[pnrobinson]

note of course that gene->protein is not always true. How does biolink deal with options?

[deepakunni3]

@pwrose

For Genome and Gene, you could do something like,

subject    edge_label    object    relation
biolink:Gene    related_to    biolink:Genome    RO:0002525

where RO:0002525 is is subsequence of from Relations Ontology.

For Gene to Protein, you can have a direct link like you mentioned,

subject    edge_label    object    relation
biolink:Gene    biolink:has_gene_product    biolink:Protein    RO:0002205

where RO:0002205 is has gene product from Relations Ontology.

[deepakunni3]

note of course that gene->protein is not always true. How does biolink deal with options?

biolink:has_gene_product can be used to link a gene to either a protein or a RNA that is transcribed from the gene.

Also, it depends on how we define the properties of a gene. Any gene that is not a protein coding gene could be represented with a qualifier that indicates as such. Biolink Model does not have such a qualifier yet.