18 November 2014 PAINT Conference Call: Difference between revisions

From GO Wiki
Jump to navigation Jump to search
 
Line 33: Line 33:
# Huaiyu: But this can still be mistaken as a lost of function, which is not what I intended.
# Huaiyu: But this can still be mistaken as a lost of function, which is not what I intended.
# Paul: PAINT is to build an evolutionary model, therefore, the curator should made the call based on the best of his knowledge. If there is not enough information, and the curator decides not to propagate, it implies that the function is lost.
# Paul: PAINT is to build an evolutionary model, therefore, the curator should made the call based on the best of his knowledge. If there is not enough information, and the curator decides not to propagate, it implies that the function is lost.
# Huaiyu: Then the user guide is not correct. Needs to be updated.
# Pascale: One can also annotate each paralog individually, so those that do not have the function won't get annotated.
# Pascale: One can also annotate each paralog individually, so those that do not have the function won't get annotated.
# Huaiyu: The problem is that the function is conserved all the way to the root of eukaryotes.
# Huaiyu: The problem is that the function is conserved all the way to the root of eukaryotes.

Latest revision as of 14:03, 19 November 2014

PAINT call Agenda and Minutes Nov. 18, 2014

Participants: Suzi, Paul T, Karen, Huaiyu, Pascale, Marc, Moni, Chris


Agenda

Annotation of families for Synapse Jamboree

Please see email for the PDF files.

  1. Huaiyu assigned all genes to Panther families
  2. Huaiyu will mark these on the Google PAINT family annotation tracking as 'Synapse jamboree' https://docs.google.com/spreadsheets/d/1uHgcaXO7t9__9GuXgBibT0H-YlnU252IsRGWIyhCWN8
  3. Karen will send Huaiyu her notes and Huaiyu will compile the information
  4. Huaiyu will make all these information available on google drive.

p.s.

  1. The location for Panther families of the gene list: https://docs.google.com/spreadsheets/d/11AHteWEti0fIal2Gt00Xm1YZFO7FielftZqNyIE0jGY/edit#gid=1657372967
  2. The location for Karen's annotation: https://docs.google.com/spreadsheets/d/1NX1YJmNTZS8HYCczob-rva8HLxBiCra4ej0HICxywTA/edit#gid=1487994982

Source(s) for annotations from PAINT tool versus Protein2GO (Karen)

Deferred to Managers Call on November 19

Use NOT and STOP (Huaiyu)

When "NOT" annotation is used in PAINT to a node, there are several scenarios. The "NOT" annotation can be applied

  • only to the term, but all parent terms are still true.
  • to the term and some parent terms, while other parent terms are still true.
  • to the term and all parent terms.

Does PAINT support this right now?

Also, according to the user guide, we can use "STOP". Is it still supported by PAINT? I have been using NOT in a lot places that I was supposed to use STOP. I am just concerned these are incorrect representations.

Minutes

  1. Huaiyu: while curating the synaptic protein (synaptotagmin, PTHR10024), there are situations that all the above NOT annotations need to be used. For example, there is term "regulation of calcium dependent exocytosis" that is conserved. When this term is propagated, its parent terms "regulation of exocytosis" and "calcium dependent exocytosis" are also propagated. However, some of the duplicated members lost the calcium dependency. Therefore, a NOT should be placed on the term as well as the parent term "calcium dependent exocytosis", but the regulation of exocytosis is still true. Currently, it doesn't seem to be possible to do it in PAINT. Also, there are a few new members that have little experiments done on them, so there is little information about them except that they maybe involved in exocytosis. There is no evidence to say whether they are involved in the process in the calcium dependent or independent way. Therefore, it is probably appropriate to put a STOP to the term "regulation of calcium dependent exocytosis".
  2. Paul: The STOP is not used in PAINT anymore. The NOT with the evidence code IRD can be used in place of stop to show a weaker evidence or not propagating the term.
  3. Huaiyu: But this can still be mistaken as a lost of function, which is not what I intended.
  4. Paul: PAINT is to build an evolutionary model, therefore, the curator should made the call based on the best of his knowledge. If there is not enough information, and the curator decides not to propagate, it implies that the function is lost.
  5. Huaiyu: Then the user guide is not correct. Needs to be updated.
  6. Pascale: One can also annotate each paralog individually, so those that do not have the function won't get annotated.
  7. Huaiyu: The problem is that the function is conserved all the way to the root of eukaryotes.
  8. Suzi: For the way to handle NOT, it is better to send out the family for people to view before the meeting. Also, in order to address this, please send the family example.
  9. Huaiyu: will send out the family, and some screenshots.