search

obophenotype / uberon

An ontology of gross anatomy covering metazoa. Works in concert with https://github.com/obophenotype/cell-ontology
http://obophenotype.github.io/uberon/
Other
131 stars 29 forks source link

Merge issues in composite ontology #422

Closed fbastian closed 3 years ago

fbastian commented 10 years ago

1) classes that are merged in the ext version, are dangling with no relations in the composite version, e.g.: http://purl.obolibrary.org/obo/ZFA_0000098, xref of http://purl.obolibrary.org/obo/UBERON_2000098

2) Incorrect merge of high-level terms:

cmungall commented 10 years ago

Check 2014-04-23, the situation is improved. There are still CARO classes but they are unique.

This process will be more robust by the time of your release

fbastian commented 10 years ago

Looks fine, thank you.

Otherwise, I have the impression that most classes that are merged, still appear as independent classes in the owl file (not in the obo file), e.g.:

Also, I noticed a developmental stage weirdly dangling at the root of the ontology: http://purl.obolibrary.org/obo/CS_0 "Carnegie stage"

cmungall commented 10 years ago

Try http://purl.obolibrary.org/obo/uberon/releases/2014-05-01/

Danglers should be gone.

The OWL is still not ideal. Lots of WBbt classes have no superclasses so they float to the top.

EHDAA2 CSs aren't mapping to HsapDv

Still need to do a tidy of the EHDAA2 xrefs in the ontology, a lot just vanished in earlier versions of EHDAA2

fbastian commented 10 years ago

Should I simply add xrefs to HsapDv.obo, or should I do it the OWL way?

Is there something to fix with WBbt, or should we provide you with potential superclasses?

cmungall commented 10 years ago

the CSs are an issue because they aren't prefixed IDs:

[Term]
id: CS01
name: CS01
namespace: carnegie_stage
is_a: CS:0  ! Carnegie stage

I thought there was a workaround, need to check. Though I think that EHDAA2 should just use this subset of HsapDv directly

cmungall commented 10 years ago

I made a request: https://github.com/obophenotype/human-developmental-anatomy-ontology/issues/10

JohnCork commented 10 years ago

Hi Chris,

I have just looked at the HsapDV.owl file which I did not know existed. It seems to cover much of what I was trying to do with CAVE ontology .

I have a few suggestions:

  1.  Change life cycle to life span  - life cycle goes from adult of reproductive age to child which would not include post menopausal females  etc.
  2.  The defn of life span (life cycle in your case) should be “ from fertilization to death” not “from prenatal stage to late adulthood”
  3.  I think life cycle stage should be a subclass of lifecycle not a superclass
  4.  There is no class for morula stage although it is mentioned in the definition of blastula stage
  5.  I think it would help to have general classes for “prenatal” and “postnatal” stages as terms like 1 month old human could be either
  6.  I just saw that there is a class for prenatal stage although it does not seem to have any sublasees ( e.g. all the Carnegie stages are prenatal stages.Also the definition "Prenatal stage that starts with fertilization and ends with the fully formed embryo.´should be changed to “ends at birth”.  The fully formed embryo is at Carnegie stage 23 (8 weeks).

Is this the form you mentioned at the meeting?

John

Dr. John Cork, Associate Professor Computer Imaging Laboratory, MEB Rm 6226 Department of Cell Biology & Anatomy, LSU Health Sciences Center, 1901 Perdido St., Box P6-2 New Orleans, LA 70112

Tel. 504 568 7177 Fax. 504 568 4392

URL: http://virtualhumanembryo.lsuhsc.eduhttp://virtualhumanembryo.lsuhsc.edu/

There are only 10 types of people, those who understand binary and those who don't.

From: Chris Mungall [mailto:notifications@github.com] Sent: Thursday, May 01, 2014 3:19 PM To: obophenotype/uberon Subject: Re: [uberon] Merge issues in composite ontology (#422)

I made a request: obophenotype/human-developmental-anatomy-ontology#10https://github.com/obophenotype/human-developmental-anatomy-ontology/issues/10

— Reply to this email directly or view it on GitHubhttps://github.com/obophenotype/uberon/issues/422#issuecomment-41951974.

cmungall commented 10 years ago

Thanks John!

  1. I agree - @fbastian and @ANiknejad, what do you think?
  2. I think this makes sense from a developmental perspective, and is what we have in uberon
  3. We treat the entire span as a stage.
  4. We should add
  5. I think we should be able to include this in the merged version
  6. I think you're right about the end stage, but it wouldn't be a subtype of carengie stage

Sorry if this is a bit cryptic can answer in more detail later

fbastian commented 10 years ago
  1. No problem, Chris, do you intend to change it in Uberon as well?
  2. When does it take place exactly? Between CS2 and CS3?
  3. Prenatal actually has 2 subclasses, HsapDv:0000002 embryonic human stage, and HsapDv:0000037 fetal stage. I changed the definition on the google code repository.
JohnCork commented 10 years ago

I am assuming #2 below refers to ‘morula’, if so, the morula is the later part of stage 2 when the cells have coalesced into a mass but the blastocystic cavity has not formed. Formation of the cavity marks the beginning of stage 3

Dr. John Cork, Associate Professor Computer Imaging Laboratory, MEB Rm 6226 Department of Cell Biology & Anatomy, LSU Health Sciences Center, 1901 Perdido St., Box P6-2 New Orleans, LA 70112

Tel. 504 568 7177 Fax. 504 568 4392

URL: http://virtualhumanembryo.lsuhsc.eduhttp://virtualhumanembryo.lsuhsc.edu/

There are only 10 types of people, those who understand binary and those who don't.

From: fbastian [mailto:notifications@github.com] Sent: Friday, May 02, 2014 7:44 AM To: obophenotype/uberon Cc: Cork, Robert Subject: Re: [uberon] Merge issues in composite ontology (#422)

  1. No problem, Chris, do you intend to change it in Uberon as well?
  2. When does it take place exactly? Between CS2 and CS3?
  3. Prenatal actually has 2 subclasses, HsapDv:0000002 embryonic human stage, and HsapDv:0000037 fetal stage. I changed the definition on the google code repository.

— Reply to this email directly or view it on GitHubhttps://github.com/obophenotype/uberon/issues/422#issuecomment-42026985.

ANiknejad commented 10 years ago

Hello,

regarding the comment of Dr. John Cork:

"4. There is no class for morula stage although it is mentioned in the definition of blastula stage" and " the morula is the later part of stage 2 when the cells have coalesced into a mass but the blastocystic cavity has not formed. Formation of the cavity marks the beginning of stage 3"

the hsapdv.obo ontology currently has this structure:

-prenatal stage --embryonic stage ---CS01 ---cleavage stage ----CS02 ---blastula stage ----CS03 ----CS04 ----CS05 ...

and follows the CS stages as described here (in fact we have to update the links in our ontology, will be done soon):

http://embryology.med.unsw.edu.au/embryology/index.php?title=Carnegie_stage_table

Stage 2 events: morula cell division with reduction in cytoplasmic volume, blastocyst formation of inner and outer cell mass

It sounds like here the 'morula stage' could be simply a synonym of CS02? Or maybe better to be more precise and to create two sub-stage of CS02? -early part of stage 2 -later part of stage 2 = morula

Thank you for your help,

Best regards,

Anne

Anne Niknejad Biocurator at http://bgee.unil.ch SIB Swiss Institute of Bioinformatics Dept Ecology & Evolution Biophore, UNIL-Sorge UNIL-CH-1015 Lausanne Switzerland

JohnCork commented 10 years ago

Hi Anne,

Although the term morula has been described as ‘not appropriate’ for mammals (http://carnegiescience.edu/publications_online/developmental_stages.pdf) it is commonly used when describing human embryology. I do have a couple of extra comments.

“It sounds like here the 'morula stage' could be simply a synonym of CS02? Or maybe better to be more precise and to create two sub-stage of CS02? -early part of stage 2 -later part of stage 2 = morula”

This would be correct. The early part of stage 2 is equivalent to “cleavage stage”

“Stage 2 events: morula cell division with reduction in cytoplasmic volume, blastocyst formation of inner and outer cell mass “

Blastocyst formation marks the boundary between stages 2 and 3, so formation of the inner and outer cell masses is not detectable until stage 3.

John

Dr. John Cork, Associate Professor Computer Imaging Laboratory, MEB Rm 6226 Department of Cell Biology & Anatomy, LSU Health Sciences Center, 1901 Perdido St., Box P6-2 New Orleans, LA 70112

Tel. 504 568 7177 Fax. 504 568 4392

URL: http://virtualhumanembryo.lsuhsc.eduhttp://virtualhumanembryo.lsuhsc.edu/

There are only 10 types of people, those who understand binary and those who don't.

From: ANiknejad [mailto:notifications@github.com] Sent: Monday, May 12, 2014 7:32 AM To: obophenotype/uberon Cc: Cork, Robert Subject: Re: [uberon] Merge issues in composite ontology (#422)

Hello,

regarding the comment of Dr. John Cork:

"4. There is no class for morula stage although it is mentioned in the definition of blastula stage" and " the morula is the later part of stage 2 when the cells have coalesced into a mass but the blastocystic cavity has not formed. Formation of the cavity marks the beginning of stage 3"

the hsapdv.obo ontology currently has this structure:

-prenatal stage --embryonic stage ---CS01 ---cleavage stage ----CS02 ---blastula stage ----CS03 ----CS04 ----CS05 ...

and follows the CS stages as described here (in fact we have to update the links in our ontology, will be done soon):

http://embryology.med.unsw.edu.au/embryology/index.php?title=Carnegie_stage_table

Thank you for your help,

Best regards,

Anne

Anne Niknejad Biocurator at http://bgee.unil.ch SIB Swiss Institute of Bioinformatics Dept Ecology & Evolution Biophore, UNIL-Sorge UNIL-CH-1015 Lausanne Switzerland

— Reply to this email directly or view it on GitHubhttps://github.com/obophenotype/uberon/issues/422#issuecomment-42825900.

ANiknejad commented 10 years ago

Thank you very much, John. So I propose to do the following changes in hsapdv.obo

-prenatal stage --embryonic stage ---CS01 ---CS02 ----cleavage stage ----morula ---blastula stage ----CS03 ----CS04 ----CS05 ... Do you agree?

JohnCork commented 10 years ago

Anne,

That structure seems fine to me. The only other thing I would mention is that the embryonic period CS1 – CS23 is sometimes divided into pre-implantation and post-implantation with implantation starting at CS 05.

John

Dr. John Cork, Associate Professor Computer Imaging Laboratory, MEB Rm 6226 Department of Cell Biology & Anatomy, LSU Health Sciences Center, 1901 Perdido St., Box P6-2 New Orleans, LA 70112

Tel. 504 568 7177 Fax. 504 568 4392

URL: http://virtualhumanembryo.lsuhsc.eduhttp://virtualhumanembryo.lsuhsc.edu/

There are only 10 types of people, those who understand binary and those who don't.

From: ANiknejad [mailto:notifications@github.com] Sent: Monday, May 12, 2014 10:34 AM To: obophenotype/uberon Cc: Cork, Robert Subject: Re: [uberon] Merge issues in composite ontology (#422)

Thank you very much, John. So I propose to do the following changes in hsapdv.obo

-prenatal stage --embryonic stage ---CS01 ---CS02 ----cleavage stage ----morula ---blastula stage ----CS03 ----CS04 ----CS05 ... Do you agree?

— Reply to this email directly or view it on GitHubhttps://github.com/obophenotype/uberon/issues/422#issuecomment-42847422.

ANiknejad commented 10 years ago

Thank you for your comments, John, I will do the changes in hsapdv.obo following and including your explanations and keep you informed.

Anne

ANiknejad commented 10 years ago

Hi Chris,

I managed to create the 'morula stage' (point 2 of your message below), thanks to John's comments (hsapdv.obo updated on /developmental-stage-ontologies.googlecode.com) Is there another point below I could manage or do you plan to do things automatically later?

Also about last comment of John: "the embryonic period CS1 – CS23 is sometimes divided into pre-implantation and post-implantation with implantation starting at CS 05."

I guess, adding one class for 'pre-implantation' and one for 'post-implantation (that should include implantation start?)' would create another level of complexity in hsapdv.obo. Carnegie stage 05 already has the definition: def: "Blastula stage during which implantation occurs." What do you think about? Thank you!

Anne

--- Message original --- Objet: Re: [uberon] Merge issues in composite ontology (#422) De: Chris Mungall notifications@github.com À: obophenotype/uberon uberon@noreply.github.com Cc: ANiknejad anne.niknejad@unil.ch Date: vendredi, 02/05/2014 01:07:56

Thanks John!

  1. I agree - @fbastian and @ANiknejad, what do you think?
  2. I think this makes sense from a developmental perspective, and is what we have in uberon
  3. We treat the entire span as a stage.
  4. We should add
  5. I think we should be able to include this in the merged version
  6. I think you're right about the end stage, but it wouldn't be a subtype of carengie stage

Sorry if this is a bit cryptic can answer in more detail later Definition of Carnegie stage 05 already talks about implantation

gouttegd commented 3 years ago

WARNING: This issue has been automatically closed because it has not been updated in more than 3 years. Please re-open it if you still need this to be addressed addressed addressed – we are now getting some resources to deal with such issues.