#11563 Parentage of GO:0000981

[rach_huntley]

There seems to be some inconsistency with the parentage of transcription factor activity terms.
In 2011 the part_of relation between GO:0003700 sequence-specific DNA binding transcription factor activity and GO:0006351 transcription, DNA-templated was removed, but there still exists a part_of relation between GO:0000981 sequence-specific DNA binding RNA polymerase II transcription factor activity and GO:0006366 transcription from RNA polymerase II promoter (see attached screenshot).

Was this overlooked or is it intentional?

Thanks,
Rachael.

[David Osumi-Sutherland]

Assigning to David Hill as I have no idea about the reasoning behind this, and he is likely to know the history.

[Karen Christie]

Hi,

The fact that

GO:0000981 sequence-specific DNA binding RNA polymerase II transcription factor activity
has part_of relationships to both:
- GO:0006366 transcription from RNA polymerase II promoter
- GO:0006357 regulation of transcription from RNA polymerase II promoter

is intentional. David and I discussed this issue of whether transcription factors are part of the process or regulate it (specifically with respect to TFIIIC) with a group of researchers in coming to this conclusion.

What looks like the error to me is that
GO:0003700 sequence-specific DNA binding transcription factor activity
does not have a part_of relationship to
- GO:0006351 transcription, DNA-templated
as well as to
- GO:0006355 regulation of transcription, DNA-templated

-Karen

[rach_huntley]

Thanks for your reply Karen,

I thought the removal of the part_of relation between GO:0003700 and GO:0006351 was intentional?

If you look here http://www.ebi.ac.uk/QuickGO/GTerm?id=GO:0003700#term=history
the relation was added in November 2010 then removed in October 2011.
I remember when it happened as it removed a whole lot of annotations as well, I even used it as an example of how small changes in the ontology can have large effects on the annotations.

Anyway, I'll leave it to you to decide which is the error!

Thanks,
Rachael.

[Karen Christie]

Hi Rachael,

Thanks for pointing out that QuickGO has a log of changes by term. That's quite handy!

My recollection is that David and I completed the transcription overhaul in summer 2011, so I think the last change I made is this one:

2011-06-09 Added RELATION part of GO:0006355 (regulation of transcription, DNA-templated)

These two subsequent changes in 2011 wouldn't have been mine:

2011-10-13 Deleted RELATION part of GO:0006351 (transcription, DNA-templated)
2011-09-04 Added SYNONYM transcription factor activity

To me, deleting the RELATION part of GO:0006351 (transcription, DNA-templated) seems like the error since David and I consciously decided that we should consider the transcription factor activity terms to be both part of "transcription" and part of "regulation of transcription".

-Karen

[David Hill]

OK. I added the relation back.

[David Hill]

Hi Rachel,

Actually Karen and I have done some rethinking about this. We think that the part_of relationship does not hold true for all types, particularly for transcription factors that negatively regulate transcription. My new plan is to remove the part_of transcription links for the generic terms and to only add the part_of transcription links to the terms that are specifically involved in positive regulation of transcription. Let me know if this doesn't make sense.

-D

[rach_huntley]

Hi David, Karen,

I think I understand what you mean, just to confirm - you are saying that any term that is a "transcription factor activity involved in positive regulation of transcription" will have part_of relationships to both positive regulation of transcription and transcription, whereas any term that is "transcription factor activity involved in negative regulation of transcription" will have a part_of relationship only to negative regulation of transcription?

I think this would be a good point to bring to the May 26th annotation call on the MGI annotation consistency paper discussions, do you think?

Thanks,
Rachael.

[David Hill]

Hi Rachael,

Yes. This is correct. So any transcription factor that positively regulates transcription is a part of the transcription process as well. All other nucleic-acid binding and protein-binding transcription factor terms are only part of the regulation of transcription.

I made the edits yesterday.

-David