TODO list

[lldelisle]

I put here what I would like to put in pygenometracks.

Far future:

• BAM track
• add the possibility to make background
• put the barcolorscale aligned below.

[DonStephano]

Hi,
is the issue with the arcs/links solved? I also suggest that the arcs should start/end in the middle of the region (maybe as an option). But I really like the improvements for blotting arcs, keep up the fantastic work!

[lldelisle]

Hi,
No sorry this is still the start of the first to the end of the second... You are right I should put an option and this would solve the issue...

[DonStephano]

No worries. It would be cool, for now I just manipulated my arc-file with the different start/end regions

[DonStephano]

It would be also very very cool to plot 4C-like plots with pygenometracks (implementation of chicPlotViewpoint) :) @joachimwolff

[joachimwolff]

It is on the list: #234

[DonStephano]

perfect! Thx a lot
great work 👍

[lldelisle]

Support UTR/coding plot for bed8 and bed9.

[lldelisle]

Propose a global parameter to be more lenient on log for negative values.

[lldelisle]

@DonStephano the links is fixed in version 3.5 with the option use_middle.

[lldelisle]

From BCC2020, +1 for BAM format.

[DonStephano]

Hey @lldelisle
thanks a lot! I will try it soon.
Additionally from a seminar where I presented pyGenomeTracks: +1 bam format :)
Thanks for the great work!

[lldelisle]

Add label of bed tracks below the regions

[GDelevoye]

Hello,

While this is just a workaround, remark that bedtools bamtobed can be used to turn a .bam into a .bed file which, in turn, is handled correctly by pyGenomeTracks. This could be either a temporary solution, or maybe even the basis of a future PR to handle the .bam files. One cannot install easily bedtools from PyPi, but unofficial python bindings exist, and bedtools can be easily installed through conda. That's probably not the best option for a clean PR, but this would just work

[lldelisle]

Thanks @GDelevoye .
My idea of a BAM track would handle some features supported by IGV: per base coverage (optionally with colors for mismatches), pair matching for PE data etc. But this is a lot of work.

[GDelevoye]

I agree, I thought the same a few days ago. This would really be a killer feature. But this is indeed very ambitious