preventing users including entire datasets (i.e contaminants and low confidence) in physical interactions

[ValWood]

Antonia pointed out instances of users including all proteins in a pull down, including "common contaminents"

We should probably change the text here:

something to say that the data should be "pre processed" to filter common false positives.

It's difficult to word because we can't use "threshold", many of the contaminants are going to be higher than the real interactors .....

ideally they would report any "conservative" dataset that they report in the paper, but if they don't do this, it's difficult to know how to word it.

suggestions?

[ValWood]

• [ ] for this particular dataset, Antonia you should contact the authors
  we need a "boiler plate" e-mail for future instances which are not caught by the annotation guidance.

Something t say: We noticed that you have reported your Mass spec dataset in it's entirety. We encourage you to remove known/likely contaminants ( translation factors, and ribosomal proteins for example)

could you draft something and circulate? @mah11 any ideas?

[mah11]

suggestions [for the interaction interface]?

I have added this to the documentation in git:

"We recommend that you only annotate interactions that you think are biologically meaningful. For example, don't include known or suspected contaminants from mass spectrometry results (e.g. ribosomal proteins, translation factors, 'sticky' proteins)."

You could crib some or all of that to put in the curation popup.

@mah11 any ideas [for email draft]?

One idea: please don't put an apostrophe in the possessive pronoun "its" ("it's" is a contraction of "it is").

Some possible words:

"In your annotations for [paper], it looks like you have included an entire set of mass spectrometry results in the physical interactions section. We believe that interaction annotations are useful when they represent biologically relevant interactions rather than experimental artefacts, so if any of the interacting proteins you have annotated are likely to be contaminants (such as ribosomal proteins or translation factors), we recommend that you remove them from the Canto annotations."

[ValWood]

Its a typo

[mah11]

now you're just taking the piss ;)

(would it be funnier if i'd typed "your"?)

[ValWood]

Yes but you just couldn't do it could you?

[mah11]

at least I thought about it!

[ValWood]

ACtion, include these suggested words

"In your annotations for [paper], it looks like you have included an entire set of mass spectrometry results in the physical interactions section. We believe that interaction annotations are useful when they represent biologically relevant interactions rather than experimental artefacts, so if any of the interacting proteins you have annotated are likely to be contaminants (such as ribosomal proteins or translation factors), we recommend that you remove them from the Canto annotations."

BUT could we add a tiny bit of text to discourage people from entering likely artifacts to start with?

[ValWood]

Maybe the message I would like needs to be further forward, before they even add the genes for the session.

here:

[ValWood]

Perhaps If you have large datasets please consider our <bulk annotation formats link https://www.pombase.org/submit-data/bulk-annotation >
Note: Only supply high confidence interactions for large datasets

@mah11 @Antonialock ?

[ValWood]

Implement text change

[kimrutherford]

Implement text change

Done!