Heatmap layout

[Piergiorge]

Hello again!

I am using phylostratr to recover data about the conservation level (and age) of analogous enzymes in the human genome, and the software is fantastic for this purpose. In our work, the results produced with BLASTp, against a customized dataset (78 proteomes - https://www.ebi.ac.uk/reference_proteomes), were processed (considering evalue, Score, Identity, and coverage), and some missing data were observed. I don't have problems with these missing data, but it would be helpful if the color of the border in the heatmap was drawn in black (the matrix will represent the evalue, the presence, and absence of orthologs for each gene). Also, species with long names showed overlap with the heatmap.

What can I do to fix this?

Thank you very much!

[arendsee]

Hi @Piergiorge , thanks for the report. Could you send me the image of the matrix (you can email it to me if it is confidential)?

I've seen the long-name problem before. I "fixed" it by modifying the generated PDF in a vector editor (Inkscape). The rest of the names are there, but just behind the map. This is not a real solution, of course. I'll label this as a bug and see if I can fix it.

I'll look into the border pattern and get back with you.

[Piergiorge]

Hello! Yes, of course! You can see that *Fusobacterium nucleatum subsp. nucleatum* ATCC 25586 and others species are behind the map. Is it possible to create shorter branches? The other thing is to change the white line around each box (I would like to change it to black). Regards,

[arendsee]

@Piergiorge Sorry if I'm missing something, but I don't see the image. Did you attach it?

[Piergiorge]

Excuse me. I made a mistake.

So ... this is the picture (I just removed the genes). In white are genomes 'without' orthologs.

[arendsee]

Cool, I think what we need is some extra wiring for passing parameters. Currently, most of them are hardcoded. I'll do a bit of refactoring.

[fka21]

Hey @arendsee,

Has this issue been resolved? I also get an oddly looking heatmap with many warning messages regarding factoring. I have uploaded a picture of the first page. Also is there a way of shifting the tree with the genes? It would feel a bit more natural at least for me.

[arendsee]

Hi @fka21

Just to clarify, you would like to be able to move the tree to the left so that the names do not overlap with the heatmap?

Could you post the warnings you are getting?

[fka21]

Hi @arendsee

I'd like to move the tree edges to be the columns and the genes to become rows. Don't know if that is more clear, I might explain it really poorly.

Also the warning messages look like this:
Warning messages: 1: In [<-.factor(tmp, is.na(ordered), value = 5) : invalid factor level, NA generated 2: attributes are not identical across measure variables; they will be dropped 3: In [<-.factor(tmp, is.na(ordered), value = 5) : invalid factor level, NA generated
And so on.