Adding jbrowse visualization

[mictadlo]

Description of feature

Hi, I found jbrowse pangenome visualization plugins. The good new is it requires the graph created by pggb. Additionally, it requires:

• Generating MAF Files with wgatools
• Spliting the MAF file into manageable chunks for visualization with mafchunk
• Generate a pseudo-BED file from the chunked MAF:
• Indexing the Pseudo-BED with tabix

I think it would be great to have the additional visualisation option to MutliQC.

Thank you for your considaration.

Michal

[subwaystation]

Hi Michal, how do you imagine this?

JBrowse is an interactive visualization tool which comes with its own code data base. So far, I integrated all the 1D and 2D visualzations as PNGs into MultiQC.

Ah, on 2nd thoughts: You would like to have these additional outputs from the pipeline so people can view it in JBrowse? That would be cool!

[subwaystation]

I had a quick look at the JBrowse way. My concern here is that it directly takes the PAF files as input. So just the pairwise alignments without taking the subsequent pangenome graph and local MSA normalizations into account. This means the alignments rely heavily on wfmash's exact base level alignment capabilities. And there was no MSA normalization of all input sequences at once.

[subwaystation]

But, this would be the first tool to actually take a look at the initial alignments. For quality control and other things. Very important I just realized.

[subwaystation]

The first step here would be to bring all required software tools to bioconda @mictadlo. Then I can add a module for each tool and integrate them into the pipeline. Could you please initiate this? Thanks!