Closed pgaudet closed 5 years ago
This is the matrix rules.
There are a set of rules for the intersections which have been checked and should be zero.
It should be a soft-check. People will either remove or let me know and I'll fix the rule.
I have forgotten where the rules live.... I can look into this next week. I don't think this rule has ever been active?
@cmungall @kltm sorry I will resurrect the manuscript too as soon as 2 others are finished... Do you remember where the rules are?
Val
To catch up people on the matrix rules, the Biocuration 2017 presentation: http://bit.ly/go-report-int-slides The report lives here: http://bit.ly/go-report-int (needs to be kicked) The rules for the report here: https://github.com/geneontology/shared-annotation-check
Just so there is no ambiguity
The code for the matrix checks are here: https://github.com/geneontology/shared-annotation-check (curators do not need to go here)
I believe Val's question was where to go for the complete set of GO rules. These live in github in the go-site repo under metadata, here: https://github.com/geneontology/go-site/tree/master/metadata/rules
But the top level doc for describing the pipeline and QC checks from a curator and provider perspective is here: http://wiki.geneontology.org/index.php/Release_Pipeline
This is the URL that should be bookmarked. The docs are not perfect and some things are still in flux but the intent is that this is the top level doc for GOC members.
@cmungall But the matrix rules themselves are at https://github.com/geneontology/shared-annotation-check/blob/master/rules.txt ; isn't this a rule for "implementing" the matrix rules?
Are the reports generated by this script available somewhere ?
https://s3.amazonaws.com/go-reports/index.html very cool! I hadn't seen that.
Maybe I can just take out the rules with lots of violations and we begin with the ones with only a few (I can quickly do all of the obvious "zero" rules).
Then I can add the big ones in slowly. Don't want to scare people....
I mena this one: https://s3.amazonaws.com/go-reports/index.html
@ValWood It's been a while, but this was part of the package we put together for Biocuration a couple of years ago :) This still needs to be "ported" to the new pipeline (https://github.com/geneontology/pipeline/issues/21), but should be easy work.
I'm sure I never saw this page though! This will come back onto my radar in a couple of months...
I've ported it; it should have a new home here from tomorrow: http://snapshot.geneontology.org/reports/shared-annotation-check.html
Talking with Seth, this is being run now every day http://snapshot.geneontology.org/reports/shared-annotation-check.html.
I will update the Rule that it's implemented and the report is at the above link.
Updated here: https://github.com/geneontology/go-site/pull/744
Is this a rule ? This is just a report that may lead to constraints and annotation review tickets, isn't it ?
Well they are 'annotation rules' but at the moment they are soft checks. I think eventually many (ideally most) will be hard checks....
@pgaudet Is the update to gorule 9 satisfactory, given the conversation?
Sure. Reports are only in 'snapshot', and not in 'current' and 'releases' ?
All reports appear in all pipeline runs, including release
(and current
, which is a copy of the latest release
).
The rules describe both things we wish and things we have, as a method of documentation and tracking progress towards our goals. Implementations and hardness/softness will change through time, as well as how something is implemented and reported. In this case, the intersection check and report predates the current pipeline reports. In the end, we hope to have all reports centralized, but for now, for some use cases, there may be some outliers.
So where do I create and edit the intersection rules? Is it a YAml file? I can't remember...
All it says right now is title: "Annotation Intersection Alerts"
@ValWood What is this expected to do ?