And maybe add a DB level constraint (if possible) on the feature_cvtermprop table so that rows the type "evidence" must have a value from the ECO cvterm.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
We need to load ECO too.
And maybe add a DB level constraint (if possible) on the feature_cvtermprop table so that rows the type "evidence" must have a value from the ECO cvterm.
Sorry I got that wrong. We consistently use the name of the evidence in Chado. We use either GO evidence names or ECO term names.
See also:
https://www.ebi.ac.uk/panda/jira/browse/PB-1591
Note to self - the latest OBO file is here:
https://code.google.com/p/evidenceontology/source/browse/trunk/eco.obo
Midori has made a mapping file:
pombe-embl/chado_load_mappings/ECO_evidence_mapping.txt
We should use that to map evidence names from flat file to ECO term IDs when storing each annotation in Chado.
ECO:0001098 'microscopy evidence'
ECO:0001098 been in the mapping file for a while now.