-
Notifications
You must be signed in to change notification settings - Fork 5
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
add male fruitless stacks #177
Comments
We have Python code for adding the individuals to the DB. Just need a ref for source + table of name + type. |
If these have not been added yet, can you provide a tsv of names, types and file paths? Otherwise please close. Cheers, David |
Where are we with this? Can we fit it into a December release. (Assuming we can so assigning to December release Milestone) |
This information (except for types) was sent to Nestor in July 2013, so I'm assuming it's somewhere in Edinburgh. I can provide it again if need be. See https://gist.github.com/mmc46/ca968041d4ef37f766a2 Files are here |
@dosumis can we start the ID's from VFB_00016227 as we have a massive gap or are we keeping this for anything? |
Individuals for these are now in the OWL. Have kept short names for now, but we should probably harmonise names with others from same dataset. (Note - having serious doubts about sustainability of keeping long, descriptive names for individuals). Committed to Sandbox, so should be visible there shortly. |
Something the ontserver doesn't like:
Caused by: org.semanticweb.owlapi.expression.ParserException: Encountered C888C3DB-AEFA-447F-BD4C-858DFE33DBE7 at line 1 column 62. Expected one of:
java.lang.NullPointerException |
The new individuals OWL does seem odd as previously: Where as the new one has just Declaration - might be further on but seems odd: |
@dosumis if you can give me a list of vfbIDs to the file names I can add them in ready. |
Looks like failure to merge in a support file. Could be an issue with shell
|
Found the problem. While trying to fix the reasoning bug, I switched out shortFormIDs with leading digits - as I suspected one problem may be a bug in shortForm parsing. Have updated individual build code to reflect this. Rebuilding now. BTW - the apparent difference in content you flagged is just a difference in line ordering. The same axioms are in the new file - just not in the same place. |
Mapping here: https://gist.github.com/dosumis/410663ada7e1778f44ad |
Still problems with OWL gen. Fixing now. |
Files added to new file system |
Definitely still not happy: Files are all live on sandbox: |
BTW Would rather solve #442 than add to VFB DB. |
Still seeing: Encountered C888C3DB-AEFA-447F-BD4C-858DFE33DBE7 at line 1 column 62. Expected one of: Class name Object property name Data property name not inverse Self But this looks fine in the file. Are you sure it's running the latest. I did post an interim version that was still broken ealier this afternoon. One thing we need to fix. The VFB build shell script is not failing when jobs in the middle of the script fail. Posting of bad files is mainly due to that (although I think not in this case). Will fix tomorrow - or you could have a go if you have time. |
Have you forced the ontology server to restart - should auto restart when owl changed plus also worth restarting tomcat to be sure - all done from jenkins. |
Still giving a similar error |
The individuals file is working OK if deployed on jenkins think you must of had a merge issue. Live on the sandbox server if you want to check everything is there. |
Deploy from Jenkins to Sandbox would be good. But we need to work on the Jenkins shell script first. Needs to be set up to return false if intermediate jobs fail. |
All ready just need to either add to VFB DB or bypass to owl - going to take a look at feasibility this weekend. |
@mmc46 do we have the FlyCircuit ids? |
Hi Robbie, I'll provide the FlyCircuit id mapping. (roll on ditching the VFB postgres DB...) |
@dosumis Could we include the remote URI in the OWL? |
Scratch that these things are from FlyCircuit - so request makes no sense. Yes, it is possible to include the remote URI in the OWL. My first instinct was to use the standard xref annotation property. The axioms have value = string that is conventionally structured as DB:accession. This is what we already use, along with a standard recipe, for rolling linkouts. We could (ab)use this field by adding values of the form (source:full_url), but this will be messy to parse. So, I'm inclined to add a new annotation property for data_source_link with string value = full URL. This can be generated from a recipe specified in the data_source table of the VFB_owl database. |
So to clarify these are from CacheroOstrovsky2010 and the external link should be http://flybase.org/reports/FBrf0211926.html or would http://flybrain.mrc-lmb.cam.ac.uk/dokuwiki/doku.php?id=pub:neuroanatomy:fru:start be more useful as the Publication would be in the definition. @dosumis can we change to FBrf (when available) rather than PMID just to be more consistent. |
The source should be http://flybase.org/reports/FBrf0211926.html. On 23 Feb 2015, at 14:19, Robbie1977 notifications@github.com wrote:
|
@mmc46 This is only for the source link - usually the link to the original data and the paper is already in the references. The wiki page does a better job of explaining the alignment and giving links to original data and hence I think is a more useful link? |
I can see that it would be useful, but possibly confusing because the wiki page gives info on the registration of the original data, not on the bridged one that we have on the site (and the only place where the template is named is in the command in the 'Registration parameters' section). Regardless, the link to the wiki could be added here. http://www.virtualflybrain.org/site/vfb_site/image_data_downloads.htm |
Then how about an internal link to the correct section of the image data downloads page. |
Ok. |
Agreed but this is only where the original data is not directly available On Mon Feb 23 2015 at 4:08:23 PM mmc46 notifications@github.com wrote:
|
Only female ones have been added, because we didn't have the male and female terms in the ontology at the time.
Now that we do (mmc46 to check it's all ok) add the stacks to the site, and to the image downloads.
The text was updated successfully, but these errors were encountered: