Very useful tool. I want to use SpaTalk in our lab. I am working on the 10x visium ST dataset (multiple samples : 2 replicate each).

I split my integrated Seurat object based on the sample id i.e A1, B1, C1 and D1 and decided to apply SpaTalk on each section one by one.

obj <- createSpaTalk(st_data = as.matrix(st_count_a1),
st_meta = st_meta_a1,
species = "Mouse",
if_st_is_sc = F,
spot_max_cell =1)

Everything works fine until this step and created SpaTalk object.

length(colnames(obj@data$rawdata))
1800
length(rownames(st_meta_a1 ) )
1800

Next I would like to use deconvolution result from cell2location which consist of weights ( spot id x cell types) .

To create input for single cell count data, I used function https://satijalab.org/seurat/reference/averageexpression to get genes_labels x celltypes matrix and saved as sc_count.

celltypes <- c("Tcells" , "Bcells", "Macro")

obj <- dec_celltype(obj,
sc_data = sc_count,
sc_celltype = celltypes,
if_doParallel = F,
use_n_cores = 3,
dec_result = st_meta_cells_fil
)

It gives error such as :

Error in dec_celltype(obj, sc_data = sc_mat1, sc_celltype = celltypes, :
Please provide a correct dec_result matrix! See demo_dec_result()!

My cell2location result matrix is in same format as required.
Tcells Bcells Macro
CA_1_1 0.07227784 0.002217575 0.003214246
CT_1_1 0.11673413 0.005541645 0.008517662
AG_1_1 0.30936356 0.007579740 0.048766695
TT_1_1 0.14501309 2.209639668 0.147608649
CC_1_1 0.24153571 0.970593423 0.296744515

But still I am having this error. Please correct me if I am doing something wrong at any step.
I would appreciate all your suggestion!!!

Thanks in advance