Annotation Conf. Call, January 27, 2015: Difference between revisions

From GO Wiki
Jump to navigation Jump to search
Line 14: Line 14:
** Marcus: Great idea. Yes, we can do this.
** Marcus: Great idea. Yes, we can do this.
* DavidH: There is an issue when we are annotating to terms using the IPI evidence code. In cases where IPI is used as a 'guilt by association' evideence code the IPI evidence code is not direct evidence. For example an annotation to a process because a proetin is know to interact with another protein that have been proven to act in that process is not direct evidence. However, for the specific GO terms that are about binding, the IPI evidence code is direct evidence for the interaction. In this sense an annotation of:<br>
* DavidH: There is an issue when we are annotating to terms using the IPI evidence code. In cases where IPI is used as a 'guilt by association' evideence code the IPI evidence code is not direct evidence. For example an annotation to a process because a proetin is know to interact with another protein that have been proven to act in that process is not direct evidence. However, for the specific GO terms that are about binding, the IPI evidence code is direct evidence for the interaction. In this sense an annotation of:<br>
Protein A 'protein binding' IPI Protein B is equivalent to an annotation
Protein A 'protein binding' IPI Protein B is equivalent to an annotation<br>
Protein A 'protein binding' IDA has_direct_input' protein B.
Protein A 'protein binding' IDA has_direct_input' protein B.<br>
**Marcus: good question, but no answer at the moment.
**Marcus: good question, but no answer at the moment.
*Birgit: use of evidence codes in complex portal?
*Birgit: use of evidence codes in complex portal?


* Judy: Is there a way to include information on algorithm used for say orthology prediction (version)? How to capture that detail? Is this something we want to include in the ECO itself?
* Judy: Is there a way to include information on algorithm used for say orthology prediction (version)? How to capture that detail? Is this something we want to include in the ECO itself?

Revision as of 14:37, 29 January 2015

Agenda

ECO presentation by Marcus Chibucos

Please post any questions you may have for Marcus on the wiki!

Discussion

  • We used Bluejeans for this conference call and it worked great.

ECO presentation

  • Slides:
  • Kimberley: Can we add references to evidence code terms. There are lot of assays reported in the papers and it will be lot of work for curators to to figure out which code to use and if the assay is represented already or not. If a reference is available for each term, then curators can check the paper and understand the scope of the code/assay.
    • Marcus: Great idea. Yes, we can do this.
  • DavidH: There is an issue when we are annotating to terms using the IPI evidence code. In cases where IPI is used as a 'guilt by association' evideence code the IPI evidence code is not direct evidence. For example an annotation to a process because a proetin is know to interact with another protein that have been proven to act in that process is not direct evidence. However, for the specific GO terms that are about binding, the IPI evidence code is direct evidence for the interaction. In this sense an annotation of:

Protein A 'protein binding' IPI Protein B is equivalent to an annotation
Protein A 'protein binding' IDA has_direct_input' protein B.

    • Marcus: good question, but no answer at the moment.
  • Birgit: use of evidence codes in complex portal?
  • Judy: Is there a way to include information on algorithm used for say orthology prediction (version)? How to capture that detail? Is this something we want to include in the ECO itself?