Creating a Seurat object with Visium public available data

Question

[charms.charms]

Hi all, I just wanted to know if any of you have experience creating a Seurat object with GSE data. I downloaded the files from GEO, and I'm getting the counts (h5 file), the scale factors (json), the image (png) and the tissue positions (csv). I've been trying to create a Seurat object, so far no success.

Thanks in advance.

[Ram]

Vissium

It's Visium with just one s. I've fixed your post.

[bk11]

Could you provide the GEO accession that you are working on?

[charms.charms]

Sure it is GSE188257

Answer 1

[bk11]

You can download dataset of GSE188257 from here. Here I m showing how to create a Seurat object for 1 dataset. You can follow the same process for other three datasets-

    #There are 4 datasets in GSE188257. I m working on only GSM5673398_655 set
        1. Perform these in your terminal.
            mkdir 01-GSM5673398_655
            cd 01-GSM5673398_655
            ls
            GSM5673398_655_counts.csv.gz              
            GSM5673398_655_scalefactors_json.json.gz    
            GSM5673398_655_tissue_positions_list.csv.gz
            GSM5673398_655_filtered_feature_bc_matrix.h5  
            GSM5673398_655_tissue_lowres_image.png

            #create a spatial directory inside 01-GSM5673398_655
            mkdir spatial
            cd spatial
           #Copy spatial data into spatial dir, unzip and rename them as in 10x output
            cp ../GSM5673398_655_scalefactors_json.json.gz .
            cp ../GSM5673398_655_tissue_* .
            gunzip *gz
            ls
            GSM5673398_655_scalefactors_json.json  
            GSM5673398_655_tissue_lowres_image.png  
            GSM5673398_655_tissue_positions_list.csv

            mv GSM5673398_655_scalefactors_json.json scalefactors_json.json
            mv GSM5673398_655_tissue_lowres_image.png tissue_lowres_image.png 
            mv GSM5673398_655_tissue_positions_list.csv tissue_positions_list.csv
            ls
            scalefactors_json.json  
            tissue_lowres_image.png  
            tissue_positions_list.csv

        2. Perform these in R    
            library(Seurat)

            data_dir="/User/01-GSM5673398_655"

            s655 <- Seurat::Load10X_Spatial(
              data.dir = data_dir, 
              filename = "GSM5673398_655_filtered_feature_bc_matrix.h5",
              assay = "Spatial", # specify name of the initial assay
              slice = "slice1", # specify name of the stored image
              filter.matrix = TRUE, 
              to.upper = FALSE
            )
            s655
            An object of class Seurat 
            32285 features across 841 samples within 1 assay 
            Active assay: Spatial (32285 features, 0 variable features)
            1 layer present: counts
            1 image present: slice1

            head([e-mail])
            orig.ident nCount_Spatial nFeature_Spatial
            AAACCGGGTAGGTACC-1 SeuratProject          26570             6507
            AAACCGTTCGTCCAGG-1 SeuratProject          30001             6605
            AAACTGCTGGCTCCAA-1 SeuratProject          45459             6943
            AAAGGCTCTCGCGCCG-1 SeuratProject          20792             5455
            AAAGGGATGTAGCAAG-1 SeuratProject          19440             5338
            AAAGTAGCATTGCTCA-1 SeuratProject          25035             6590

[charms.charms]

Thank you so much for your help, it was really helpful.

Answer 2

[Nat.Nataren]

Have you looked at the Seurat vignettes for guidance?

Edit: this vignette in particular goes through analysing Visium data

https://satijalab.org/seurat/articles/spatial_vignette

[charms.charms]

Thanks you so much !

[leviathan]

this vignette does not explain how to load data from files. does 10X have a tutorial for this?

[Nat.Nataren]

This the page for the Load10X_Spatial() function, which bk11 used in his reply above (if you refer to step 2 in his answer). https://satijalab.org/seurat/reference/load10x_spatial

Also, once you are more familiar with the Seurat workflows, the command cheat sheet is a nice reference as well :) I often still refer to this, as I can never remember all the functions and their arguments! https://satijalab.org/seurat/articles/essential_commands