trgt plot feature plans

[bw2]

Hi, I'm thinking about how to best generate read visualizations for genotypes in TRExplorer.

The current TRGT plots are helpful, but don't have several useful features of previous visualization tools that let you see non-reference bases, repeat counts, etc.:

The TRGT plot images don't currently specify substituted bases, repeat units, or repeat counts:

Are there plans to add such features to TRGT plots?

Is deepdive meant to replace/supercede the plot functionality?

What is the meaning of this blue/gray stripes at the top of the image?

Thanks,
-Ben

[tmokveld]

Hi Ben,

I had a quick stab at implementing nucleotide-level resolution functionality (for consensus and mismatches):

This works reasonably well for small expansions, but since we are not constrained by short reads, we can resolve much larger repeat expansions. For these longer alleles, per-nucleotide annotations become impractical, as the bases must shrink to the point where they are barely visible. At that scale, a higher-level summary or compressed representation becomes necessary to preserve interpretability:

Still, it is worth thinking more about this, and I would be happy to hear any suggestions on how such visualizations could scale without losing clarity or interpretability.

Regarding the gray stripes; these indicate regions within the tandem repeat that do not match any of the motif definitions (see the documentation for more details). When nucleotides are shown, this becomes nicely illustrated:

[bw2]

Thanks Tom! and sorry I somehow missed that docs page.
I think the change you made is a good one. Making the read background color lighter and flatter (ie. without the 3D effect) might further improve readability.
As far as handling different scales, it seems reasonable to have some length threshold beyond which you revert to the original visualization without base letters, maybe still using color rectangles to specify nucleotides until a second threshold is reached (the same way that IGV behaves when you zoom out).

[bw2]

An even more informative visualization would segment the repeats by constituent motifs - like this plot from Mahreen Khan's poster at ASHG:

[tmokveld]

Thanks for your suggestions! I'll spend some time further working this out.

[wdecoster]

I happened to stumble upon this thread, and maybe things like the visualization suggested above are already solved with my aSTRonaut script, see also here. Feedback appreciated!

[bw2]

Cool visualizations @ https://github.com/wdecoster/pathSTR . Thanks for the pointer @wdecoster

I just prototyped some changes to the trgt plot subcommand to test out the visual features we discussed above @
https://github.com/bw2/trgt/blob/main/README.md

It needs testing on different types of repeat loci, but seems to work ok on basic examples.

[bw2]

After some further updates, this is what the plot currently looks like for a large CNBP expansion: