Issue with plotEmbedding with GeneExpressionMatrix

[jmmuncie]

Hi - I've encountered an issue using plotEmbedding with GeneExpressionMatrix to visualize expression of genes of interest in UMAP space on my dataset of integrated scATAC and scRNA data from a 10x Multiome experiment. The function works as expected for most of my genes of interest, but for a few examples, I get a blank plot, as seen below (with log file attached).

It seems that the genes that don't work have relatively few non-zero entries in the gene expression matrix, e.g. a few hundred out of 37,000+ cells. Is there a minimum number of non-zero entries required for the plotEmbedding function to work? Or do you think something else might be causing the issue here?

ArchR-plotEmbedding-43a34d14939b-Date-2022-06-05_Time-11-49-04.log

[rcorces]

Hi @jmmuncie! Thanks for using ArchR! Please make sure that your post belongs in the Issues section. Only bugs and error reports belong in the Issues section. Usage questions and feature requests should be posted in the Discussions section, not in Issues.

Before we help you, you must respond to the following questions unless your original post already contained this information:
1. If you've encountered an error, have you already searched previous Issues to make sure that this hasn't already been solved?
2. Can you recapitulate your error using the tutorial code and dataset? If so, provide a reproducible example.
3. Did you post your log file? If not, add it now.
4. Remove any screenshots that contain text and instead copy and paste the text using markdown's codeblock syntax (three consecutive backticks). You can do this by editing your original post.

[rcorces]

I'd guess that this is related to the quantile cut that is performed to remove outlier values from the plot. see the quantCut parameter. If you set this to quantCut = NULL, I would imagine that it should display the few cells with non-zero values.

[jmmuncie]

Yes, that sounds like it should work. However, when I set quantCut = NULL as in the following:

plotEmbedding(
    ArchRProj = proj_Mef2c_v13_E9,
    colorBy = "GeneExpressionMatrix",
    name = "Sox17",
    embedding = "UMAP_Harmony_Combined",
    quantCut = NULL
)

I get an error message:

Plotting Embedding
1 Error in quantile.default(x, probs = c(lo, hi)) : 'probs' outside [0,1]
In addition: Warning messages:
1: In min(quantCut) : no non-missing arguments to min; returning Inf
2: In max(quantCut) : no non-missing arguments to max; returning -Inf

And traceback() yields this:

8: stop("'probs' outside [0,1]")
7: quantile.default(x, probs = c(lo, hi))
6: quantile(x, probs = c(lo, hi))
5: quantile(x, probs = c(lo, hi))
4: .quantileCut(plotParamsx$color, min(quantCut), max(quantCut))
3: FUN(X[[i]], ...)
2: lapply(seq_along(colorList), function(x) {
       if (getArchRVerbose()) 
           message(x, " ", appendLF = FALSE)
       plotParamsx <- .mergeParams(colorList[[x]], plotParams)
       if (plotParamsx$discrete) {
           plotParamsx$color <- paste0(plotParamsx$color)
       }
       if (!plotParamsx$discrete) {
           plotParamsx$color <- .quantileCut(plotParamsx$color, 
               min(quantCut), max(quantCut))
           plotParamsx$pal <- paletteContinuous(set = plotParamsx$continuousSet)
           if (!is.null(pal)) {
               plotParamsx$pal <- pal
           }
           if (is.null(plotAs)) {
               plotAs <- "hexplot"
           }
           if (!is.null(log2Norm)) {
               if (log2Norm) {
                   plotParamsx$color <- log2(plotParamsx$color + 
    ...
1: plotEmbedding(ArchRProj = proj_Mef2c_v13_E9, colorBy = "GeneExpressionMatrix", 
       name = "Sox17", embedding = "UMAP_Harmony_Combined", quantCut = NULL)

ArchR-plotEmbedding-5e203dec417f-Date-2022-06-06_Time-10-31-28.log

[rcorces]

ok. that looks like a bug. I will look into that. in the mean time, I think quantCut = c(0,1) should do the same thing.

[rcorces]

After looking into this, I dont think your problem is related to quantCut. That part of the function should behave as expected because it has a built-in check to see if the upper-quant limit is 0 and if so it ignores that upper limit. I think you problem is more likely due to 1 of two things:

1. you havent set plotAs so it defaults to hex and creates an aggregate signal for all cells falling in that hex bin
   or
2. Even if you try plotAs = "points" I cant remember how ArchR sets the order of points plotted but it probably is not smart enough to show the few hundred colored cells on top so they might get hidden behind a sea of grey points

I'll still fix the quantCut = NULL bug though

[rcorces]

The quantCut = NULL bug should be fixed via cadcd31 on release_1.0.2

Feel free to close if the comments on hex and points above fix your problem. Otherwise, happy to help troubleshoot more.

[jmmuncie]

Thanks as always for the help! quantCut = c(0,1) did the trick for me. Obviously not a lot of useful information in the plot, but when I zoom in, I can tell it's properly plotting values now.

However, quantCut = NULL still throws the following error and does not produce any plot.

Plotting Embedding
1 Error in quantile.default(x, probs = c(lo, hi)) : 'probs' outside [0,1]
In addition: Warning messages:
1: In min(quantCut) : no non-missing arguments to min; returning Inf
2: In max(quantCut) : no non-missing arguments to max; returning -Inf

[rcorces]

I do not get an error with quantCut = NULL. Perhaps your ArchR install wasnt properly loaded after updating to the latest release_1.0.2.

Safest way to make sure is to install using devtools, then run:

#to unload a package and reload
detach("package:ArchR", unload=TRUE)
library(ArchR)

Here is what my output looks like

> p_PAX5 <- plotEmbedding(
+   ArchRProj = projHeme5,
+   colorBy = "GeneScoreMatrix",
+   name = c("PAX5"),
+   embedding = "UMAP",
+   quantCut = NULL,
+   imputeWeights = getImputeWeights(ArchRProj = projHeme5),
+   plotAs = "points"
+ )
Getting ImputeWeights
No imputeWeights found, returning NULL
ArchR logging to : ArchRLogs/ArchR-plotEmbedding-816a455eb4bdd-Date-2022-06-06_Time-12-51-07.log
If there is an issue, please report to github with logFile!
2022-06-06 12:51:09 : Getting Matrix Values...

Plotting Embedding
1 
ArchR logging successful to : ArchRLogs/ArchR-plotEmbedding-816a455eb4bdd-Date-2022-06-06_Time-12-51-07.log

[jmmuncie]

Yeah, that was all it was. Thanks again!