Error with IntegrateLayers

[JoyOtten]

Hi,

It seems that I can't use the IntegrateLayers function. I have two layers in my object however I'm getting the following error codes:

test <- IntegrateLayers(
object = seurat_soup_doublet_remove_mito, method = CCAIntegration,
orig.reduction = "pca", new.reduction = "integrated.cca", assay = "SCT",
verbose = FALSE
)
| 0 % ~calculating Error in UseMethod(generic = "Assays", object = object) :
no applicable method for 'Assays' applied to an object of class "NULL"

seurat_soup_doublet_remove_mito <- IntegrateLayers(
object = seurat_soup_doublet_remove_mito, method = HarmonyIntegration,
orig.reduction = "pca", new.reduction = "harmony",
verbose = FALSE)

Error in validObject(object = x) :
invalid class “Seurat” object: 1: all cells in assays must be in the same order as the Seurat object
invalid class “Seurat” object: 2: all cells in assays must be in the same order as the Seurat object
invalid class “Seurat” object: 3: all cells in reductions must be in the same order as the Seurat object
invalid class “Seurat” object: 4: all cells in reductions must be in the same order as the Seurat object
invalid class “Seurat” object: 5: all cells in graphs must be in the same order as the Seurat object (offending: SCT_nn)
invalid class “Seurat” object: 6: all cells in graphs must be in the same order as the Seurat object (offending: SCT_snn)
In addition: Warning messages:
1: In harmony::HarmonyMatrix(data_mat = Embeddings(object = orig), :
HarmonyMatrix is deprecated and will be removed in the future from the API in the future
2: Quick-TRANSfer stage steps exceeded maximum (= 4536250)

In all honesty, I have no idea what goes wrong over here. I will attach an image on how my seurat object looks like.
Before integration, I filtered the data, SCTransform, RunPCA, RunUMAP, FindNeighbors, FindClusters.

[wangbenwang123]

If you need to perform IntegrateLayers, you must create a new assay split by sample or treatment before running SCTransform. The code is as follows:

sc[['Integrated']] <- split(sc[['RNA']], f = sc$sample) # Split based on sample type in the metadata

DefaultAssay(sc) <- 'Integrated'

sc <- SCTransform(sc, assay = 'Integrated') # SCTransform defaults to using 'RNA' if the assay is not specified, so you need to set 'Integrated' here.

DefaultAssay(sc) # After SCTransform, the default assay is set to 'SCT', so I recommend explicitly specifying the assay in all function parameters if you're unsure about the default behavior.

DefaultAssay(sc) <- 'SCT'

sc <- RunPCA(sc, npcs = 30, verbose = F, assay = 'SCT')

options(future.globals.maxSize = 3e+09) # Adjust based on your actual needs

sc <- IntegrateLayers(

  object = sc,

  method = RPCAIntegration, # Or you can use CCA or Harmony as the method

  normalization.method = "SCT",

  verbose = F,

  assay = 'SCT'
)

I hope this helps resolve your issue.

[JoyOtten]

Hi,

I already tried that the problem is that my layers are already split.
! The following layers are already split: 'counts.1', 'counts.2' Please join before splitting
Run rlang::last_trace() to see where the error occurred.

From the start I had two seurat objects since they are two independent replicates from one another.

[wangbenwang123]

你好

我已经尝试过了，问题是我的图层已经被分割了。 ！以下图层已被拆分： 'counts.1'， 'counts.2' 请在拆分 Run 之前加入以查看错误发生的位置。rlang::last_trace()

从一开始，我就有两个 seurat 对象，因为它们是两个彼此独立的复制品。

Please follow my code order and parameters strictly as instructed.

[JoyOtten]

I can't read chinese characters and the code above I can't follow since my Seurat object is from the start already split.

[maggietsui]

Having the same no applicable method for 'Assays' applied to an object of class "NULL" error using RPCA, although I am using standard log normalization and not SCT.

[mhkowalski]

Hi,

It seems like there is something going wrong with the structure of your object, as I haven't encountered this issue before. Could you please provide the code on how you're constructing seurat_soup_doublet_remove_mito from your two individual seurat objects?