neurogenomics / EpiCompare

Comparison, benchmarking & QC of epigenetic datasets
https://doi.org/doi:10.18129/B9.bioc.EpiCompare
12 stars 3 forks source link

Percentage overlap: bar plot #128

Closed bschilder closed 1 year ago

bschilder commented 1 year ago

For input in bed6 format (without the additional 4 columns), only one color is visible in the provided example after excluding the last 4 columns. Reference itself may be excluded.

Screenshot 2023-01-31 at 16 31 46

This may be do to a hard-coded fill scale somewhere. Could you investigate @serachoi1230 ?

bschilder commented 1 year ago

@serachoi1230 can you tell me which function you altered to do this?

I noticed the plot you included in the reviewer comments reply is rather different from the one EpiCompare is generating by default:

In reviewer comments reply

Screenshot 2023-02-14 at 20 18 56

In report

tmp

Am I looking at the wrong plot, or does the plot in the comments just need to be updated?

bschilder commented 1 year ago

Ok, so this comes from the fact that there's a difference between how the BED formats are handled.

I've also modified the code so that the adjusted p-values are rounded to the nearest 4 significant digits in either case. adb82337-0b33-47d8-aefa-ebb24d890f22

I've also made both versions interactive, and conditionally add the labels to avoid the warning from plotly about not being able to transfer over these kinds of geoms: Screenshot 2023-02-16 at 14 36 18

bschilder commented 1 year ago

I've also renamed p.adjust --> q-value, and flipped the x/y axes to be more consistent across the plots generated for different BED formats. 6fa1c724-1fb1-467a-8be3-4775d346cff1