search

yhoogstrate / fuma

:dash::leopard: FuMa: reporting overlap in RNA-seq detected fusion genes
GNU General Public License v3.0
5 stars 8 forks source link

Make FuMa work with Python3 #48

Closed Redmar-van-den-Berg closed 3 years ago

Redmar-van-den-Berg commented 3 years ago

I've updated FuMa to work with python3 (https://github.com/yhoogstrate/fuma/issues/46). All tests work, although I had to change some of the ordering for the fusion results in the test cases, since the ordering is different in Python3. To be honest, I'm not entirely sure why this is, as far as I can see the ordering as it is now should be correct, e.g. chr1 comes before chr10. However, I'm not sure if the ordering of the fusion results is relevant for the functionality of FuMa, so please have a critical look at my changes. This is the first time I've attempted to convert a project from Python2 to Python3.

Redmar-van-den-Berg commented 3 years ago

Hi @yhoogstrate I figured moving FuMa to Python3 might solve those Conda dependency issues, since I'm not having any luck figuring out what is going wrong there. This is a pretty big change to FuMa, so I don't want to merge this myself. Can you have a look at this PR at some point? No rush.

yhoogstrate commented 3 years ago

@Redmar-van-den-Berg OMG!! Thanks! Busy this week with teaching and tight deadlines but I will look at it afterwards. If all tests pass I'm fine with you merging it (but please bump to 4.* and update the Changelog). The changes are not that radical.

Also thanks for the cleanups of my horrible non python convention style code.

 + from Fusion import AD_DIRECTION_REVERSE
 + from Fusion import AD_DIRECTION_FORWARD
 - from .Fusion import AD_DIRECTION_REVERSE
 - from .Fusion import AD_DIRECTION_FORWARD

I remember those dots were a hack to get the functional test working.