Suggested additional Panache features

[brettChapman]

I've had a few users working with Barley Panache now. A request I have is an ability to highlight selected genes of interest, perhaps a tick box to show only genes being sorted by PAV so that it becomes easier to use a panel of genes in a single view for publication figures. Also an export/reporting feature of the selected gene regions to show PAV of that region, and the genes present with their percentage PAV. These would improve the functionality of Panache a lot. Thanks.

[SingingMeerkat]

Hi @brettChapman , Thank you for the new dedicated GitHub issue. Please let me know if the following could cover your needs or is a correct understanding of what you would like to achieve:

1. Highlighting selected genes of interest: At a specific region, you would like to be able to hide visually some genes from the annotation track while keeping on display only some selected by the user. We could imagine a list containing the names of the genes present in the current window, with an option to toggle their visibility individually. This would hide/show their glyphs on the annotation track accordingly. It is also possible to highlight them by adding color or some kind of shadow behind them instead of hiding them completely. Highlighting by displaying them with a special color would have my preference. We could even imagine enabling users to attribute a color for each gene of their liking, though that would be more difficult to implement. A possible middle ground would be to enable a choice of color between the default black, a highlight color, and background grey (which would be equivalent to hiding them). When you say "being sorted by PAV", do you mean that the list of genes to highlight should be sorted with the most present gene on top for example? To make it easier to select the genes that should be highlighted?

2. Export features You would like to extract a file crossing the information of PAV on a region with the gff information, something that could look like this:

   # Panache PAV x GFF export
   # PAV: pavFile.name
   # Genes: GffFile.name
   # Region: CHROMNAME from START to STOP
   GeneID       |   %PAV   | StartPos | StopPos | Genome1 |  Genome2 ...
   someGene     |   100%   |  25448   |  36547  |    1    |    1
   anotherGene  |    97%   |    362   |  16695  |    0    |    1

   ... sorted with the most present gene on top.

[SingingMeerkat]

Hi @brettChapman ,

I took some time to add an export function button that you can find in the latest versions of Panache, added with commits 7644990 (for the bananaGenomeHub branch) and commit d81a3d7 (for the main branch). As long as both a pav file and a gff file are available, a download button is visible:

It enables users to export a file looking like the following: '# Panache PAV x GFF export '# PAV: NameOfYourPavFile.pav '# Genes: NameOfYourGffFile.gff '# Region: chr01 from 17597524 to 17832564	AnnotID	mean	start	stop	A	B	C	D	E	F	G	H	I	J	K	L	M	N	O
Ma01_g20380	86.67	17685364	17688075	100	100	100	0	100	0	100	100	100	100	100	100	100	100	100
Ma01_g20410	86.67	17777074	17777936	100	100	100	100	100	0	100	100	100	0	100	100	100	100	100
Ma01_g20340	80	17620823	17621956	100	100	100	0	0	0	100	100	100	100	100	100	100	100	100

In Panache, when you click on the download button, it creates this 'coverage' matrix, where for each annotation ID it computes a 'coverage' score. This score per genome is based on the proportion of the annotation that is covered by at least one present panBlock. A mean score is calculated and all annotations are sorted, with the most covered/present on top. The downloaded file extracts only the annotation that are visible on screen when pressing the button, but the data for the whole chromosome are created anyway so we could download these if you prefer.

I have just noticed that the header is not writing properly, I will be working on it right away.

[SingingMeerkat]

Done! I corrected the matrix so that it displays this header now: annotID meanCoverage annotStart annotStop A B C D E

[brettChapman]

Thanks @SingingMeerkat

Sorry for the delay, I've been preoccupied with other projects. I'll test out the new updated and provide feedback when I can.

Where should the GFF and PAV files be located? At the moment I just have the corresponding JSON files uploaded to Panache

[brettChapman]

Hi @SingingMeerkat

I tried with the new commit 'https://github.com/SouthGreenPlatform/panache/commit/76449907eb5600a578748bd4a097632ab5f3d7cc` on the bananaGenomeHub branch.

git cloned the branch and then did a git checkout 7644990

I noticed while ediitng the LocalFilter.vue file that new entries are available: this.updatePavFileName('OriginOfPavData.pav'); this.updateGffFileName('OriginOfGffData.gff');

Should these point to the PAV file and GFF file used to generate the JSON files, and should they be in the Public folder? and should they have the exact same naming convention as the JSON files?

I did all that, and found that the page doesn't load up.

[SingingMeerkat]

Hi @brettChapman , sorry for the delay, I started a PostDoc a month ago and did not have much time to check that yet.

About this.updatePavFileName('OriginOfPavData.pav'); and this.updateGffFileName('OriginOfGffData.gff');

These two lines ask the app to store the file names of the .pav and .gff that had been used to create the .json file that is preloaded, so that when you are using the app you can see on which file/data you are working.

Normally, you could put whatever you want in it as it is supposed for display purpose only.

It should not prevent the page to load though, I'll look into it. I had troubles with the package-lock.json file (a configuration file for managing the required packages) recently, that could be related.

[brettChapman]

Hi @SingingMeerkat

I managed to get it working somehow. I have found the excel spreadsheet doesn't match up to the coverage reported though for the genes. I'm looking at gene HORVU.MOREX.r3.2HG0208140. The excel spreadhsheet says either 100% or 0%. While the hover over feature for the search criteria within Panache shows some genomes only have like 80% coverage.