How to Plot Multiple Modules onto one Violin Plot (VlnPlot())

[ksaunders73]

From #3521 I learned how to plot multiple genes onto one plot using AddModuleScore():

tnbclist <- list(c("FABP7", "TSPAN8", "CYP4Z1", "HOXA10", "CLDN1",
                 "TMSB15A", "C10ORF10", "TRPV6", "HOXA9", "ATP13A4",
                 "GLYATL2", "RP11-48O20.4", "DYRK3", "MUCL1", "ID4", "FGFR2",
                 "SHOX2", "Z83851.1", "CD82", "COL6A1", "KRT23", "GCHFR",
                 "PRICKLE1", "GCNT2", "KHDRBS3", "SIPA1L2", "LMO4", "TFAP2B",
                 "SLC43A3", "FURIN", "ELF5", "C1ORF116", "ADD3", "EFNA3",
                 "EFCAB4A", "LTF", "LRRC31", "ARL4C", "GPNMB", "VIM", 
                 "SDR16C5", "RHOV", "PXDC1", "MALL", "YAP1", "A2ML1",
                 "RP1-257A7.5", "RP11-353N4.6", "ZBTB18", "CTD-2314B22.3", "GALNT3",
                 "BCL11A", "CXADR", "SSFA2", "ADM", "GUCY1A3", "GSTP1",
                 "ADCK3", "SLC25A37", "SFRP1", "PRNP", "DEGS1", "RP11-110G21.2",
                 "AL589743.1", "ATF3", "SIVA1", "TACSTD2", "HEBP2"))
sobj <- AddModuleScore(object = sobj, features = tnbclist, name = "TNBC_List") 
VlnPlot(sobj, features = "TNBC_List1")

From this I am able to plot one module onto a single violin plot, with the x axis being separated by cluster. Is there a way to plot multiple modules on the same plot, and not just the two plots listed side by side.

For example, to have the x axis be this "TNBC_List1" and this "ER_List1"?

ERlist <- list(c("CPB1", "RP11-53O19.1", "TFF1", "MB", "ANKRD30B",
            "LINC00173", "DSCAM-AS1", "IGHG1", "SERPINA5", "ESR1",
            "ILRP2", "IGLC3", "CA12", "RP11-64B16.2", "SLC7A2",
            "AFF3", "IGFBP4", "GSTM3", "ANKRD30A", "GSTT1", "GSTM1",
            "AC026806.2", "C19ORF33", "STC2", "HSPB8", "RPL29P11",
            "FBP1", "AGR3", "TCEAL1", "CYP4B1", "SYT1", "COX6C",
            "MT1E", "SYTL2", "THSD4", "IFI6", "K1AA1467", "SLC39A6",
            "ABCD3", "SERPINA3", "DEGS2", "ERLIN2", "HEBP1", "BCL2",
            "TCEAL3", "PPT1", "SLC7A8", "RP11-96D1.10", "H4C8",
            "PI15", "PLPP5", "PLAAT4", "GALNT6", "IL6ST", "MYC",
            "BST2", "RP11-658F2.8", "MRPS30", "MAPT", "AMFR", "TCEAL4",
            "MED13L", "ISG15", "NDUFC2", "TIMP3", "RP13-39P12.3", "PARD68"))

sobj <- AddModuleScore(object = sobj, features = ERlist, name = "ER_List") 

I tried a suggestion from this issue #3366 but unfortunately did not work (assuming I did this correctly), and just placed the two plots side by side instead of on one single plot:

sobj1 <- sobj
sobj1[['TNBC']] <- CreateAssayObject(data = t(x = FetchData(object = sobj1, vars = 'TNBC_List1')))
sobj1[['ER']] <- CreateAssayObject(data = t(x = FetchData(object = sobj1, vars = 'ER_List1')))
VlnPlot(sobj1, features = c('TNBC_List1', 'ER_List1'))

Hopefully this makes sense, and thanks for reading!

[saketkc]

While this is currently not supported, you can pull the values using FetchData and then plot these using ggplot2:
Example:

> mydata <- FetchData(object = pbmc_small, vars = c('nCount_RNA',  'nFeature_RNA', 'ident'))
> mydata
               nCount_RNA nFeature_RNA ident
ATGCCAGAACGACT         70           47     0
CATGGCCTGTGCAT         85           52     0
GAACCTGATGAACC         87           50     0
TGACTGGATTCTCA        127           56     0
AGTCAGACTGCACA        173           53     0

You would melt the data frame to make it long form and assign nCount_RNA and nFeature_RNA as variables which you could then use to color by in ggplot.

[ksaunders73]

Thank you very much, @saketkc !

I think this is what you were saying I should do (but correct me if I am wrong):

sobjlists <- FetchData(object = sobj, vars = c("ER_List1", "TNBC_List1"))
head(sobjlists)

library(reshape2)
melt(sobjlists, id.vars = c("ER_List1", "TNBC_List1"))

vln <- ggplot(sobjlists, aes(x= ER_List1, y = TNBC_List1, color = ER_List1)) + geom_violin()
vln

Is there a way to have this plotted per cluster or is it only able to be compared to each other do you think?

[saketkc]

You would also need also idents say the "seurat_clusters" column in FetchData which would be your xaxis. Since this is more of a ggplot2 question, I would encourage you to explore other forums.

[ksaunders73]

Thank you so much again @saketkc!

Due to my data having basically every cell in that seurat_clusters column, and the x-axis as a result having way too many entries, I tried this instead, essentially making a new column in the meta data listing the cluster identity, and then making that column the x axis. I then colored each violin plot by my gene list:

#after labeling my UMAP
sobj[["ClusterName"]] <- Idents(object = sobj)

sobjlists <- FetchData(object = sobj, vars = c("ER_List1", "TNBC_List1", "ClusterName"))
library(reshape2)
melt(sobjlists, id.vars = c("ER_List1", "TNBC_List1", "ClusterName"))
p <- ggplot() + geom_violin(data = sobjlists, aes(x= ClusterName, y = ER_List1, fill = ER_List1, colour = "ER+ Signature"))+ geom_violin(data = sobjlists, aes(x= ClusterName, y = TNBC_List1, fill = TNBC_List1, colour="TNBC Signature"))

See this page for lots of customization tips: http://www.sthda.com/english/wiki/ggplot2-violin-plot-quick-start-guide-r-software-and-data-visualization

Thanks again @saketkc and I hope this all helps someone else with their work!