runBayesSpaceClustering

[schak-ukf]

When I tried to run runBayesSpaceClustering, I encountered the following error: "spot_df" not found!

Here is the error:

run the pipeline

spata_obj <-

• runBayesSpaceClustering(

• object = spata_obj, 
  

• name = "bayes_space", # the name of the output grouping variable
  

• qs = 5
  

• )
  Error: object 'spot_df' not found

Spata 2.0 version is: 3.1.1 and Seurat Version is 5.0. Can you please recommend some steps to overcome the error.?

[wuxiangning]

Hello, have you resolved this issue? I recently encountered the exact same problem.

[kueckelj]

Hello, can you reinstall the package and see if the error persists?

[wuxiangning]

Hello!Should I directly download this new version you updated today? Or maybe I can simply update the existing package without uninstalling and reinstalling it?

[wuxiangning]

Hello!
I reinstalled it, but it still requires reinstalling the batch and another model package in order to start fitting. It seems that this will successfully enable Bayesian clustering.
Here r my codes:
devtools::install_github("theMILOlab/SPATA2")
devtools::install_github("theMILOlab/SPATA2","dev")
packageVersion("SPATA2")
install.packages("xgboost")
object <-

• runBayesSpaceClustering(

• object = object,
  

• name = "bayes_space", # the name of the output grouping variable
  

• qs = 5
  

• )
  16:59:11 Running BayesSpace::spatialPreprocess().
  17:01:19 Using input for q: 5.
  17:01:19 Running BayesSpace::spatialCluster().
  Neighbors were identified for 4733 out of 4733 spots.
  Fitting model...

[wuxiangning]

However, when I reinstalled to version 3.1.2, I encountered new issues.
I found that I could not directly invoke asSPATA2 to convert the Seurat object as I did with version 3.0.1. Additionally, if I try to eliminate this bug by reinstalling with devtools::install_github("theMILOlab/SPATA2", "dev"), my SPATA2 automatically reverts back to version 3.0.1. This seems to be an unsolvable problem.
packageVersion("SPATA2")
[1] ‘3.1.2’
object<- asSPATA2(

• object = seurat_object,
• sample_name = "P9T_Spatial",
• platform = "VisiumSmall",
• img_name = "image",
• img_scale_fct = "hires",
• assay_name = "Spatial",
• assay_modality = "gene",
• verbose = TRUE
• )
  17:49:05 Initiating SPATA2 object of spatial platform: VisiumSmall
  17:49:05 Transferring data.
  Error in (function (classes, fdef, mtable) :
  函数‘asMolecularAssay’标签‘"Assay"’找不到继承方法

[kueckelj]

Hello, unfortunately I can not reproduce the error you mentioned with v3.1.2. Maybe it has something to do with some specifics about your object I am not aware of. How did you create the Seurat object? I assume with Seurat::Load10X_Spatial()? Do you have multiple assays stored in the Seurat object? I am asking, because you are specifying assay_name which is not necessary if you have only one assay. Regardless, can you provide some more details about the Seurat object?

[wuxiangning]

Hello!
Since my spatial transcriptomics file is an RDS file rather than the traditional H5 file, I actually loaded it using the RDS reading method. My Seurat object also contains only one assay. Here is my loading code and the detailed components of the Seurat object:
Code：
library(SPATA2)
library(SeuratData)
P9T_Spatial <- readRDS("F:\数据分析\Rdata\原始data\JH空转data\空间转录组\P9T_Spatial.rds.gz")
SpatialFeaturePlot(P9T_Spatial, features = "nCount_Spatial") + theme(legend.position = "right")

SpatialFeaturePlot(P9T_Spatial, features = "nCount_Spatial") + theme(legend.position = "right")
spata_obj <- asSPATA2(

• object = P9T_Spatial,
• sample_name = "P9T_Spatial",
• platform = "VisiumSmall",
• img_name = "image",
• img_scale_fct = "hires",
• assay_name = "Spatial",
• assay_modality = "gene",
• verbose = TRUE
• )
  21:38:32 Initiating SPATA2 object of spatial platform: VisiumSmall
  21:38:32 Transferring data.
  21:38:34 Active assay: 'gene'.
  21:38:34 No directory was specified to store the image. Unloading won't be possible. Set with setImageDir().
  21:38:34 Active image: 'image'.
  21:38:34 Computing pixel scale factor.
  21:38:41 Done.
  警告信息:
  1: In .local(object, ...) :
  Assay 'Spatial' is in outdated Seurat format and will be converted to Seurat v5.
  2: In value[3L] : Could not find or transfer TSNE data.

[kueckelj]

As I mentioned in issue #194 it would be helpful if you can somehow provide the Seurat object since I can not reproduce the issues. Is that possible?

[kueckelj]

Can you open a new clean R session, install SPATA2 with devtools::install_github("kueckelj/SPATA2"), load SPATA2 with library(SPATA2), and see if the problem persists?