BUG | do_FeaturePlot | diverging.palette argument is not working about scpubr

[enblacar]

Hmmm they look ok, apart from the fact that now the feature to plot is "signatrue" instead of "signature"...

It seems like a proper bug. I will add it to the to-do list for the next update. I am sorry it did not work!

Many thanks for the quick feedback and feel free to open another Issue if you face other problems!

Best,
Enrique

[enblacar]

Hi @venkan,

Thanks for using my package!

In order for you to map a diverging palette to the plots, you need to use enforce_symmetry = TRUE. Otherwise, I only allow the use of sequential palettes for variables that are not centered, transformation that enforce_symmetry will perform on the data to plot if you set it to TRUE.

You can then set symmetry.type to absolute to make it centered around 0 or centered alongisde with symmetry.center to tailor the center to your liking.

For this, you need to update to v2.0.2.

Hope this helps!
Enrique

[venkan]

Thank you. When I use

SCpubr::do_FeaturePlot(scdata, features = c("signature"), reduction = "umap", split.by = "DiseaseGroup", ncol=4, order = FALSE, font.size = 8, plot.axes = TRUE, legend.length = 8, legend.width = 0.5, legend.position = "bottom", pt.size=0.3, enforce_symmetry=TRUE, diverging.palette="RdBu")

I'm using ‘2.0.2.9000’ package version. I want the scaling from -0.5 to -0.1 and centered with -0.3 with diverging.palette="RdBu". So, I used it like below:

SCpubr::do_FeaturePlot(scdata, features = c("signature"), reduction = "umap", split.by = "DiseaseGroup", ncol=4,order = FALSE, font.size = 8, plot.axes = TRUE, legend.length = 8, legend.width = 0.5, legend.position = "bottom", pt.size=0.3, diverging.palette="RdBu", symmetry.type = "centered", symmetry.center=-0.3)

But why the colors are not changed in the output plot below?

I want the colors that are used for enforce_symmetry=TRUE with the scaling I want.

[enblacar]

In your second input, there is not enforce_symmetry = TRUE. Can you try with that and let me know if it works?

Best,
Enrique

[venkan]

Thanks for your quick reply.

SCpubr::do_FeaturePlot(scdata, features = c("signature"), reduction = "umap", split.by = "DiseaseGroup", ncol=4,order = FALSE, font.size = 8, plot.axes = TRUE,legend.length = 8, legend.width = 0.5, legend.position = "bottom", pt.size=0.3, diverging.palette="RdBu", symmetry.type = "centered", symmetry.center=-0.3, enforce_symmetry=TRUE)

Output:

You see the above image. Scaling is wrong.

[enblacar]

Hmmm true:

Can you check one more thing? Try adding to what you already have:
min.cutoff = -0.5
max.cutoff = -0.1

If it does not work, I will need to have check to fix the bug!

Let me know if it works and thanks for the help!
Enrique

[venkan]

Thanks again. but didn't work.

SCpubr::do_FeaturePlot(scdata, features = c("signature"), reduction = "umap", split.by = "DiseaseGroup", ncol=4,order = FALSE, font.size = 8, plot.axes = TRUE,legend.length = 8, legend.width = 0.5, legend.position = "bottom", pt.size=0.3, diverging.palette="RdBu", symmetry.type = "centered", symmetry.center=-0.3, enforce_symmetry=TRUE, min.cutoff = -0.5, max.cutoff = -0.1)

Output:

Please check whether the arguments I gave above are right. Thank you !!

[venkan]

sorry. that's a typo (signature it is). Only when I use split.by have this problem.