Chord plots error about scpubr HOT 3 CLOSED

Hi @rbutleriii,

Thanks for using my package!

The error seems to arise due to recent implementations in liana. With version 0.1.6 the code you provided works perfectly.
I assume this is related to the changes in how interactions are sorted depending on different parameters.

I will have to go back to the code and update it to match the newest version of liana. Sorry for the inconveniences!

You can download a version of liana compatible with the function here.

Will update this once I fix it!
Best,
Enrique

from scpubr.

@enblacar this relates to my other issue, perhaps having SCpubr as a dependency of liana, and hence doing my unittests with it would resolve any such potential inconsistencies :)

from scpubr.

Hi @rbutleriii,

Sorry for taking that long in implementing the changes!

I have updated the code for SCpubr::do_LigandReceptorPlot() to work perfectly with liana v0.1.12 and will roll out in the next, big, SCpubr v2.0.0 update.

Here is an example output with your same code, adapted to the various modifications added to the function:

library(magrittr)
library(liana)
library(SCpubr)

liana_path <- system.file(package="liana")
testdata <- readRDS(file.path(liana_path , "testdata", "input", "testdata.rds"))
liana_test <- liana::liana_wrap(testdata) %>% liana::liana_aggregate()

# Chord diagram of the total significant interactions from each cluster and all the rest.
out <- do_LigandReceptorPlot(liana_output = liana_test, 
                             top_interactions = 25, 
                             return_interactions = TRUE)

# Retrieve Plot.
out$Plot

And now, you need to plot the chord diagrams yourself with the output tibbles:

## Total number of interactions per group.
SCpubr::do_ChordDiagramPlot(from_df = TRUE, 
                            df = out$`Group Interactions`, 
                            link.border.color = "black", 
                            z_index = TRUE)

## Total number of interactions per LR pair.
SCpubr::do_ChordDiagramPlot(from_df = TRUE, 
                            df = out$`LR Interactions`, 
                            link.border.color = "black", 
                            z_index = TRUE)

I hope this helps! Stay tuned for the update on CRAN! If you want to use it right now, please be aware that downloading the latest commits effectively downloads a v2.0.0 of SCpubr with all the breaking changes this involves.

Best,
Enrique

from scpubr.