error can't run escape::performNormalization?

[Sa753]

Dear team,

I have seurat object and I used the gene set provided as an example in the package. I then run runEscape and it runs well but I can't perform the normalisation because performNormalization function gives an error.

Here is the code I used

pbmc_small <- runEscape(pbmc_small,
method = "ssGSEA",
gene.sets = escape.gene.sets,
groups = 1000, min.size = 3, new.assay.name = "escape.ssGSEA")

pbmc_small <- escape::performNormalization(input.data = pbmc_small,
assay = "escape.ssGSEA",
gene.sets = escape.gene.sets)

Error in (function (cond) :
error in evaluating the argument 'x' in selecting a method for function 't': subscript out of bounds

(function (cond)
.Internal(C_tryCatchHelper(addr, 1L, cond)))(structure(list(message = "subscript out of bounds",
call = GetAssayData(object = x)[i, j, ..., drop = FALSE],
object = structure(c(80.1955263737123, -15.8737762524596, ...
5.
[.Assay(sc[[name]], "data")
4.
sc[[name]]["data"]
3.
Matrix::t(sc[[name]]["data"])
2.
.pull.Enrich(input.data, assay) at #24
1.
escape::performNormalization(input.data = pbmc_small, assay = "escape.ssGSEA",
gene.sets = escape.gene.sets)

I am using R version >4.4.2, Seurat version 4.4.0, Seuratobject 4.1.4, escape 2.5.5

Please advise
Thank you

[ncborcherding]

Hey @Sa753

Apologies for the delay, I am just getting back from traveling.

I can't recreate the issue from my end using Seurat 5.3.1. Based on the error messages, my guess is that it is a version issue in handling the sparse matrices. It has been increasingly difficult to support both version 4 and 5 for Seurat and I will likely need to start requiring a minimum Seurat version going forward.

My reproducible example/output

library(escape)
library(Seurat)

data("escape.gene.sets", package="escape")
pbmc_small <- runEscape(pbmc_small,
                        method = "ssGSEA",
                        gene.sets = escape.gene.sets,
                        groups = 1000, min.size = 3, new.assay.name = "escape.ssGSEA")

pbmc_small <- escape::performNormalization(input.data = pbmc_small,
                                           assay = "escape.ssGSEA",
                                           gene.sets = escape.gene.sets)

Computing expressed-gene counts per cell...
Normalizing enrichment scores...

Apologies as this issue will likely not be fixed, but performNormalization() is not completely necessary for downstream analysis - it can be useful for some visualizations and statistical testing. An alternative approach is to normalize the data yourself by dividing the enrichment scores by the number of counts in your meta data.

Nick