Keep filtered genes

[jasonshiller]

Description of feature

Store genes that are filtered out somewhere in the pipeline. These could be useful for development and may be of interest in certain research contexts

[rosscrowhurst]

What would your start point be? Would it be the BRAKER3 predictions final or would it be interim predictions that do not even make it to final? Would it be final (from all sources e.g. including lift over) vs the absolute final set (e.g. after running evigene, after culling some models manually etc). So what do you see as the start point and what do you see as the end point as gffcompare will do the rest once those points are defined. Also why would you want to store fragments masquerading as full models for example? Just asking

[jasonshiller]

It isn't only fragments that will be getting filtered out, agreed people are unlikely to want true fragments but there may be valid genes in there or there may be pseudogenised genes that may be of research interest. Additionally it would be useful for deciding filtering parameters e.g if you have it on full intron support for liftoff and braker genes what genes are getting thrown away (In some cases if you have not that much evidence maybe a more lenient approach is beneficial)

I haven't thought through exactly which stage/stages was just more of a reminder that I'd like to revisit this at some time later.

[GallVp]

We can try to dump a lot of files and stats if the user includes the debug profile under the -profile parameter. I have created a project issue (#96) to track these suggestions.