You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Main README and docs do not mention funannotate2 #1187
Please link to it with a brief explanation of the relative status of v1 and v2 - my impression from https://funannotate2.readthedocs.io/en/latest/changelog.html is that funannotate2 not yet production ready, but is expected to replace v1 at some point?
I'm maintaining 1.x and have fixed some bugs and things with EVM and PASA that are folded into those dependencies now that are customized fixes of those packages + some rust reimplementation speedups. Fun1 is part of my lab's workflows so I plan to maintain it for a while and have been working on getting singularity and docker images that are clean and efficient lately on a couple of OSes.
@nextgenusfs is developing 2.x - there are different predictors and consensus model calling
I have not done any direct performance and accuracy testing between the two though have been testing a lot in comparison of 1.7.x and 1.9.x and working on some streamlining I am doing in 1.10
So you'll have to decide what your preferences are for these. I've depended on the training strategies we employ in 1.x so I'm sticking with it for now.
Activity
nextgenusfs commentedon Jul 31, 2026
Funannotate2 is fully functional. Was completely rewritten for dependency reduction. I’ve switched over awhile ago. They aren’t the same.
peterjc commentedon Jul 31, 2026
Ah, looking at https://github.com/nextgenusfs/funannotate2/releases I see now that https://funannotate2.readthedocs.io/en/latest/changelog.html is totally out of date.
[Update: I logged https://github.com/nextgenusfs/funannotate2/issues/80 for that].
Still, my point stands. The v1 main documentation could and perhaps at this point should suggest looking at v2.
nextgenusfs commentedon Aug 1, 2026
Sure. Valid point. I just don’t have a lot of time. This stuff all outside of day job.
peterjc commentedon Aug 3, 2026
See also @hyphaltip's reply on nextgenusfs/funannotate2#80 (comment)
I've opened nextgenusfs/funannotate2#81 to clarify the f2 changelog page (which wrongly gave me the impression f2 was still pre-release).
Should I suggest an edit to the f1 docs that f2 be used now in preference?
Merge pull request #1234 from nextgenusfs/readme-funannotate2-note
hyphaltip commentedon Oct 12, 2026
addressed
Merge pull request #1235 from nextgenusfs/readme-funannotate2-note-ma…
rzelle-lallemand commentedon Oct 29, 2026
@hyphaltip, I was surprised to see a new release for funannotate 1.x (https://github.com/nextgenusfs/funannotate/releases/tag/v1.9.0). Do you still encourage users to migrate to 2.x? It sounds like you're planning maintain both versions for the foreseeable future?
hyphaltip commentedon Oct 29, 2026
I'm maintaining 1.x and have fixed some bugs and things with EVM and PASA that are folded into those dependencies now that are customized fixes of those packages + some rust reimplementation speedups. Fun1 is part of my lab's workflows so I plan to maintain it for a while and have been working on getting singularity and docker images that are clean and efficient lately on a couple of OSes.
@nextgenusfs is developing 2.x - there are different predictors and consensus model calling
I have not done any direct performance and accuracy testing between the two though have been testing a lot in comparison of 1.7.x and 1.9.x and working on some streamlining I am doing in 1.10
So you'll have to decide what your preferences are for these. I've depended on the training strategies we employ in 1.x so I'm sticking with it for now.