Need more information about how to deal multi-Seurat-samples

[River366]

Hi there, first let me extend my thanks for developing this useful package! But I feel some confused about how to use the package, would you mind provide more information about how to deal with multi-samples Seurat List (like the fig) before integration. I believe that will make convenience for all users!

Thank you for your guidance.

[renjun0324]

@River366
Thank you for using MURP.
Just murping the normalized data in seurat and rebuild the seurat object. My test code is as follows:

library(Seurat)
library(MURP)
library(dplyr)
data(sdata)

# 1. seurat object for test
obj_1 = CreateSeuratObject(t(sdata))
obj_2 = CreateSeuratObject(t(sdata))

# 2. normalize data
obj_1 = NormalizeData(obj_1)
obj_2 = NormalizeData(obj_2)

# 3. murp
murp_1 = MURP(Data = as.matrix(obj_1@assays$RNA@data) %>% t, fast = TRUE)
murp_2 = MURP(Data = as.matrix(obj_2@assays$RNA@data) %>% t, fast = TRUE)

# 4. create new seurat object
new_obj_1 = CreateSeuratObject(murp_1$Recommended_K_cl$centers)
new_obj_2 = CreateSeuratObject(murp_2$Recommended_K_cl$centers)

# 5. combine...

But cell labels in individual samples will not correspond after murp. I assign a new label to each murp by calculating the ratio.

[River366]

Thanks for you help!
Happy Chinese New Year!