Create slim for public release

[cmungall]

Need to check with @hkawaji - what are the criteria for being in the slim?

[hkawaji]

• The slim version should include only the samples in the phase1 freeze as leaves
  ( https://fantom5-collaboration.gsc.riken.jp/files/data/shared/FREEZE_PHASE1.1/ )
• It would be better to take out (original) internal nodes connected to only the deleted leaves

Does this sound clear?

[cmungall]

pre-release available for verification

Navigate to:
https://github.com/cmungall/fantom5-ontology/blob/master/src/ontology/ff-phase1.obo

click on 'raw'

(alternatively, git clone the repo)

This consists of

• only 1351 samples tagged phase1
• only descendant sample grouping classes (FF:0nnnnnn) of those 1351, and uberon/cl/do/etc terms that are in this graph closure

[hkawaji]

Cool. We are going to have a look. I'll get back to you soon.

[hkawaji]

Hi, Chris,

I got a repot below from my colleague, Kasukawa-san. He found the terms below are not included in the OBO file:

PATO:0002099 (which is a parent node of FF:0000245 and FF:0000246)
UBERON:0001272 (which is a parent node of UBERON:0000981)
UBERON:0009778 (which is a parent node of UBERON:0002048)

Do you think these have to be added?

best,
Kawaji H

[cmungall]

I just made a new release of the subset

• PATO:0002099 is included
• UBERON:0001272 ! innominate bone is not - but UBERON:0000981 ! femur is included. The relationship between these is one of articulation, to make the subset I only include parthood, developmental and subclass relationships

[hkawaji]

Hi,

Kasukawa-san found FF terms that are not linked to any Phase 1 samples seem to be re-appeared.

For example, FF:0000640 "Fibroblast - Aortic Adventitial, cytoplasmic fraction"
The samples linked to this term including "FF:14310-155C8" are Phase 2 samples.

[cmungall]

You're right, sorry about this, thanks for checking.

I was using the OWLAPI code to extract the graph closure, this appears too liberal. I will use my own code which I have vetted to do this and regenerate.

On Jul 11, 2013, at 6:27 PM, hkawaji wrote:

Hi,

Kasukawa-san found FF terms that are not linked to any Phase 1 samples seem to be re-appeared.

For example, FF:0000640 "Fibroblast - Aortic Adventitial, cytoplasmic fraction"
The samples linked to this term including "FF:14310-155C8" are Phase 2 samples.

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

[hkawaji]

It looks fine. I will start to make it public soon.

[cmungall]

Fantastic! Is it OK if I send a copy to the ENCODE DCC folks?

On Wed, Jul 24, 2013 at 6:04 PM, hkawaji notifications@github.com wrote:

It looks fine. I will start to make it public soon.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/6#issuecomment-21527357
.

[hkawaji]

Year, I am now thinking to prepare this directory:

http://fantom.gsc.riken.jp/5/suppl/pre_release/
(not available yet.)

Here is the current idea of the content:

https://fantom5-collaboration.gsc.riken.jp/webdav/home/kawaji/110419-FF-ontology/pre_release/

Could you please let me know your opinion (in particular, readme file).
We can move this forward if you are OK.

best,
kawaji H

[cmungall]

OK - I made a minor change in the ontology header. Ready to go!