curate drugs

[nlwashington]

need to map all the drugs in use to CHEBI, PubChem, or DrugBank.

some "drugs" are actually not specific drugs, rather classes of drugs (perhaps based on targets, or pathways, etc.) you may need to try to find relevant ontologies for those; or we may need to instantiate a version of term genie that will allow you to create those kinds of drug-groupings. if annotations to drug classes are extensive, we should consider this.

[micheldumontier]

PubChem should have all chemical entities - so i think it should act as the reference set.

[nicolevasilevsky]

Look for CAS#

[nicolevasilevsky]

@nlwashington and @kshefchek: I finished the curation of the drugs but there were several higher level classes such as AKT inhibitors. let me know if you'd like to meet to discuss.

[nlwashington]

MEK inhibitors: http://www.ebi.ac.uk/chebi/searchId.do?chebiId=CHEBI:79091

[nicolevasilevsky]

finished the second pass, added more mappings. Please let me know if you'd like to review

[micheldumontier]

yes, how to review?

m

On Apr 21, 2015, at 6:01 PM, nicolevasilevsky notifications@github.com wrote:

finished the second pass, added more mappings. Please let me know if you'd like to review

— Reply to this email directly or view it on GitHub.

[nlwashington]

@micheldumontier you can review the mapping file in git

[nlwashington]

wait, @kshefchek or @nicolevasilevsky, what's the difference between the drug-map and therapeutic_context mapping files?

[nicolevasilevsky]

@nlwashington @kshefchek, hm, not sure. I was just working on the therapeutic_context mapping file. Would you like me to add all the mappings to the drug-map file too?

Unrelated- Nicole, could you tell me how to hyperlink the file names as you did above?

[nicolevasilevsky]

@micheldumontier, I just noticed your comment about PubChem. I didn't find all the chemicals in PubChem, so some of them are mapped to Chebi, DrugBank and one is mapped to the CAS #. However, I did not search PubChem first, so some of the terms could be in PubChem, but are mapped to Chebi or DrugBank instead.

[kshefchek]

@nlwashington - therapeutic_context is a direct dump of the drug column and includes drug combinations. drug-map is a unique list of drug labels that appear in that list.

@nicolevasilevsky for consistency I mapped the drug with the CAS # to pubchem (see the drug-map file)

[nicolevasilevsky]

@kshefchek Thanks Kent! I won't touch the drug map file then.

[micheldumontier]

@nicolevasilevsky https://github.com/nicolevasilevsky i think i'm missing the context for this - where do these terms come from?

@kshefchek https://github.com/kshefchek where are the drug combinations coming from?

Michel Dumontier Associate Professor of Medicine (Biomedical Informatics), Stanford University Chair, W3C Semantic Web for Health Care and the Life Sciences Interest Group http://dumontierlab.com

On Wed, Apr 22, 2015 at 9:08 AM, nicolevasilevsky notifications@github.com wrote:

@kshefchek https://github.com/kshefchek Thanks Kent! I won't touch the drug map https://github.com/monarch-initiative/mckb/blob/master/resources/mappings/drug-map.tsv file then.

— Reply to this email directly or view it on GitHub https://github.com/monarch-initiative/mckb/issues/4#issuecomment-95250190 .

[kshefchek]

We have a small dataset of patients/patient populations with a variant and disease diagnosis that have a response to a drug or drug combination. The database is in our resources directory g2p.sql.gz if you would like to explore.

Right now we're converting the relational data to triples, although there is an on going discussion here on how to best do this: #9

[micheldumontier]

@kshefchek got it.