Calculating average mito.percentage for each cluster (seurat)

Question

I have a tricky data set with cells that will have a higher percent of mitochondria genes than "typical" data sets. I would like to look at the mito percentage in each cluster without any filtering (e.g., subset(seurat_object, subset = nFeature_RNA > 200 & nFeature_RNA < 3000 & percent.mito < 10))

FeaturePlot(seurat_object, features=c('percent.mito'), cols = c("gold", "firebrick4"), pt.size=0.5, label=TRUE, repel=TRUE)

Done!

I would also like to calculate the average percent.mito for each cluster and place the value on the FeaturePlot/TSNE plots. I have no idea how to manipulate a Seurat object to calculate these values but I can probably figure out how to place those values on the graph -- probably with geom_text.

My Hail Mary approach is below:

p3 <- FeaturePlot(seurat_object, features=c('percent.mito'), cols = c("gold", "firebrick4"), pt.size=0.5, label=TRUE, repel=TRUE)
p3 <- lapply(X = p3, FUN = function(x) mean(p3))
CombinePlots(plots = p3)

My attempts yield the error below; I don't know if my approach is close to correct and therefore, I have spent little time working through the error message.

In mean.default(p3) : argument is not numeric or logical: returning NA
Execution halted

and

p3 <- FeaturePlot(seurat_object, features=c('percent.mito'), cols = c("gold", "firebrick4"), pt.size=0.5, label=TRUE, repel=TRUE)
p3 <- lapply(X = p3, FUN = function(x) mean(seurat_object$meta.data$seurat_clusters))
CombinePlots(plots = p3)

Error in `[[.Seurat`(x, i, drop = TRUE) :
  Cannot find 'meta.data' in this Seurat object
Calls: lapply ... lapply -> FUN -> mean -> $ -> $.Seurat -> [[ -> [[.Seurat

and

p3 <- FeaturePlot(seurat_object, features=c('percent.mito'), cols = c("gold", "firebrick4"), pt.size=0.5, label=TRUE, repel=TRUE)
p3 <- lapply(X = seurat_object, FUN = function(x) mean(seurat_object$meta.data$seurat_clusters))
CombinePlots(plots = p3)

Error in as.list.default(X) : 
  no method for coercing this S4 class to a vector
Calls: lapply -> lapply -> as.list -> as.list.default```

If anyone can provide some guidance -- you will be inducted into my personal hall of fame. :)

Thank you for your time.

P.S. I am using Seurat 3.1

Update:

md = surate_object@meta.data 
md_mean <- ddply(md, .(seurat_clusters), summarize, mean_mito_per=mean(percent.mito))

This was not the approach I was hoping for but I have the values -- which when expanded to the length of the original df, I can insert it into a free slot in the seurat_object.

If anyone has a better approach I would love to hear it.

UPDATE: Using @PPK's approach

# First the cluster annotation and the tsne embeddings are merged
label.df <- cbind(as.data.frame(seurat_object@meta.data), as.data.frame(seurat_object@reductions$tsne@cell.embeddings)) 

# using dplyr across to calculate the mean mitochondrial percentage and 
# the median tsne values per cluster
label.df <- label.df %>% dplyr::group_by(seurat_clusters) %>% dplyr::summarise(dplyr::across(percent.mito, ~ mean(.), .names = "mean_{.col}"), dplyr::across(contains("tSNE"), ~ median(.)))

# running FeaturePlot and adding the label layer
FeaturePlot(seurat_object, reduction = "tsne", features = "percent.mito") + 
  geom_text(data = label.df, aes(x = tSNE_1, y = tSNE_2, label = mean_percent.mito))

Error:

* object '.col' is a function.
ℹ Input `..1` is `dplyr::across(percent.mito, ~mean(.), .names = "mean_{.col}")`.
ℹ The error occured in group 1: seurat_clusters = "0".
Backtrace:
     █
  1. └─global::initial_clustering(BM0H, "BM0H")
  2.   └─`%>%`(...)
  3.     ├─base::withVisible(eval(quote(`_fseq`(`_lhs`)), env, env))
  4.     └─base::eval(quote(`_fseq`(`_lhs`)), env, env)
  5.       └─base::eval(quote(`_fseq`(`_lhs`)), env, env)
  6.         └─`_fseq`(`_lhs`)
  7.           └─magrittr::freduce(value, `_function_list`)
  8.             ├─base::withVisible(function_list[[k]](value))
  9.             └─function_list[[k]](value)
 10.               ├─dplyr::summarise(...)
 11.               └

Answer 1

You can extract the necessary values and add them directly the plot as a second layer using plot + geom_text(). This is very similar to the inner workings of the DimPlot function with label = TRUE but allows you to use anything as label.

# I use dplyr v1.0.2 for all data frame manipulations 
# First the cluster annotation and the tsne embeddings are merged
label.df <- cbind(as.data.frame(seurat_object@meta.data), as.data.frame(seurat_object@reductions$tsne@cell.embeddings)) 

# using dplyr across to calculate the mean mitochondrial percentage and 
# the median tsne values per cluster
label.df <- label.df %>% dplyr::group_by(seurat_clusters) %>% dplyr::summarise(dplyr::across(percent.mito, ~ mean(.), .names = "mean_{.col}"), dplyr::across(contains("tSNE"), ~ median(.)))

# running FeaturePlot and adding the label layer
FeaturePlot(seurat_object, reduction = "tsne", features = "percent.mito") + 
  geom_text(data = label.df, aes(x = tSNE_1, y = tSNE_2, label = mean_percent.mito))

The labels are positioned at the median position of the cells of a cluster. Typing LabelClusters (without brackets) in the R console will show you the labeling function used by Seurat that contains some additional ways of positioning the labels.
If you want to avoid overlapping labels you can use geom_text_repel from the ggrepel package.