HOME  ›   pipelines
If your question is not answered here, please email us at:  ${email.software}

10x Genomics
Visium Spatial Gene Expression

Running spaceranger count (FFPE)

Learning objectives

In this tutorial you will:

Prerequisites

For successful run of this tutorial, you must:

Overview of Space Ranger's count pipeline

Spatial gene expression for formalin fixed paraffin embedded (FFPE) tissue is determined using spaceranger count pipeline which takes microscope image of the visium slide (in either TIFF or JPEG formats) and sample FASTQ files as inputs. The pipeline performs alignment, tissue and fiducial detection as well as barcode/UMI counting. Outputs capture the feature-spot matrices, clustering and differential gene expression (DGE) which can be further analyzed and visualized in Loupe Browser.

Example FFPE dataset

In this tutorial, we will run spaceranger count pipeline on a mouse brain FFPE section public dataset. Key dataset features are:

Download example FASTQs and reference data

All the following commands will be run in the working directory spaceranger_tutorial that was used to set up spaceranger on a compatible compute platform.

Example FFPE dataset

Both the raw sequencing files in FASTQ format, and the image in JPG format, are available for batch download on the dataset page. For better organization, we will create a datasets folder prior to downloading the required file.

# Create datasets folder
mkdir datasets
 
# Download FASTQ to datasets folder
curl https://cf.10xgenomics.com/samples/spatial-exp/1.3.0/Visium_FFPE_Mouse_Brain/Visium_FFPE_Mouse_Brain_fastqs.tar -o datasets/Visium_FFPE_Mouse_Brain_fastqs.tar
 
# Download image file to datasets folder
curl https://cf.10xgenomics.com/samples/spatial-exp/1.3.0/Visium_FFPE_Mouse_Brain/Visium_FFPE_Mouse_Brain_image.jpg -o datasets/Visium_FFPE_Mouse_Brain_image.jpg
% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
  5 4154M    5  218M    0     0  32.7M      0  0:02:06  0:00:06  0:02:00 33.0M

Alternatively, you can also use wget to download.

# Create datasets folder
mkdir datasets
 
# Download FASTQ to datasets folder
wget -P datasets/ https://cf.10xgenomics.com/samples/spatial-exp/1.3.0/Visium_FFPE_Mouse_Brain/Visium_FFPE_Mouse_Brain_fastqs.tar
 
# Download image file to datasets folder
wget -P datasets/ https://cf.10xgenomics.com/samples/spatial-exp/1.3.0/Visium_FFPE_Mouse_Brain/Visium_FFPE_Mouse_Brain_image.jpg
Resolving cf.10xgenomics.com (cf.10xgenomics.com)... 104.18.0.173, 104.18.1.173, 2606:4700::6812:1ad, ...
Connecting to cf.10xgenomics.com (cf.10xgenomics.com)|104.18.0.173|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 4356188160 (4.1G) [application/x-tar]
Saving to: ‘datasets/Visium_FFPE_Mouse_Brain_fastqs.tar’
 
37% [=======================>                                        ] 1,649,129,803  207MB/s  eta 13s    

Reference transcriptome

Since the example dataset is based on mouse tissue section, we can download the latest version of the mouse transcriptome reference available from the Downloads page. Here the curl download option is highlighted.

# Download mouse reference
curl -O https://cf.10xgenomics.com/supp/spatial-exp/refdata-gex-mm10-2020-A.tar.gz
% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
  1 9835M    1  158M    0     0  34.1M      0  0:04:48  0:00:04  0:04:44 34.1M

Probe set reference

Since this is a FFPE tissue sample, the assay uses a pair of oligonucleotide probes targeting protein coding genes. In addition to the reference transcriptome, spaceranger also requires the species specific probe set reference file in CSV format to enable analysis of FFPE samples. You can either download the probe set reference from the 10x support website or use the probe set references pre-bundled in Space Ranger.

# Download mouse probe set reference from support website
curl -O https://cf.10xgenomics.com/samples/spatial-exp/1.3.0/Visium_FFPE_Mouse_Brain/Visium_FFPE_Mouse_Brain_probe_set.csv
# Space Ranger 2.0 comes bundled with probe set files
## Source mouse probe set reference
~/spaceranger_tutorial/spaceranger-2.0.0/probe_sets/Visium_Mouse_Transcriptome_Probe_Set_v1.0_mm10-2020-A.csv
% Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 2257k  100 2257k    0     0  5579k      0 --:--:-- --:--:-- --:--:-- 5587k

In this tutorial, we will use the path associated with the first option.

Extract files

After successful download of the all the required files, the contents of the tar files need to be extracted before moving onto the next steps.

# Extract sample FASTQ files
tar -xvf datasets/Visium_FFPE_Mouse_Brain_fastqs.tar -C datasets/ && rm datasets/Visium_FFPE_Mouse_Brain_fastqs.tar
 
# Extract mouse reference transcriptome
tar -xzvf refdata-gex-mm10-2020-A.tar.gz && rm refdata-gex-mm10-2020-A.tar.gz
# Sample FASTQ files
Visium_FFPE_Mouse_Brain_fastqs/
Visium_FFPE_Mouse_Brain_fastqs/Visium_FFPE_Mouse_Brain_S3_L002_R1_001.fastq.gz
Visium_FFPE_Mouse_Brain_fastqs/Visium_FFPE_Mouse_Brain_S3_L001_I2_001.fastq.gz
Visium_FFPE_Mouse_Brain_fastqs/Visium_FFPE_Mouse_Brain_S3_L002_R2_001.fastq.gz
...
 
# Reference mouse transcriptome
refdata-gex-mm10-2020-A/
refdata-gex-mm10-2020-A/fasta/
refdata-gex-mm10-2020-A/fasta/genome.fa
...

Successful extraction will create two additional folders, highlighted in bold, within the working directory.

spaceranger_tutorial
├── datasets
│   ├── Visium_FFPE_Mouse_Brain
│   └── Visium_FFPE_Mouse_Brain_image.jpg
├── refdata-gex-mm10-2020-A
├── Visium_Mouse_Transcriptome_Probe_Set_v1.0_mm10-2020-A.csv
└── spaceranger-2.0.0

Set up the command for spaceranger count

We now have all the required inputs needed to run the spaceranger count pipeline. To obtain more information about the specifying inputs, print the pipeline specific usage statement.

# Print count usage statement
spaceranger count --help
spaceranger-count
Count gene expression and feature barcoding reads from a single capture area

USAGE: spaceranger count [FLAGS] [OPTIONS] --id <ID> --transcriptome <PATH> --fastqs <PATH>... <--image <IMG>|--darkimage <IMG>...|--colorizedimage <IMG>>   FLAGS: --no-bam Do not generate a bam file --nosecondary Disable secondary analysis, e.g. clustering. Optional --disable-ui Do not serve the web UI --noexit Keep web UI running after pipestance completes or fails --nopreflight Skip preflight checks -h, --help Prints help information ...   OPTIONS: --id <ID> A unique run id and output folder name [a-zA-Z0-9_-]+ --description <TEXT> Sample description to embed in output files --image <IMG> Single H&E brightfield image in either TIFF or JPG format --slide <TEXT> Visium slide serial number, for example 'V10J25-015' --area <TEXT> Visium area identifier, for example 'A1' --transcriptome <PATH> Path of folder containing 10x-compatible reference
...

Input image reorientation

By default spaceranger expects the hourglass shaped corner fiducial to be in the top left corner.


The input image (top right) is however rotated 90°. To ensure good fiducial alignment and tissue spots detection, it is important to correct for this shift in orientation. There are three ways to achieve this:

In this tutorial, we will choose the second option of using the --reorient-images=true flag for consistency in code from previous version of spaceranger

We can now build the spaceranger count command for the example FFPE dataset. We will running the pipeline in our working directory spaceranger_tutorial assuming the same directory structure as shown previously. The input folder paths below reflect this choice.

In case you have a different setup, amend the paths accordingly prior to running the pipeline to avoid any errors. The easiest method to customize would be to copy the code below in any text editor of your choice (e.g. notepad++), edit and paste it back to the terminal.

spaceranger count --id="Visium_FFPE_Mouse_Brain" \
                 --description="Adult Mouse Brain (FFPE) using Mouse WTA Probe Set" \
                 --transcriptome=refdata-gex-mm10-2020-A \
                 --probe-set=Visium_FFPE_Mouse_Brain_probe_set.csv \
                 --fastqs=datasets/Visium_FFPE_Mouse_Brain_fastqs \
                 --image=datasets/Visium_FFPE_Mouse_Brain_image.jpg \
                 --slide=V11J26-127 \
                 --area=B1 \
                 --reorient-images=true \
                 --localcores=16 \
                 --localmem=128
For compute platforms connected to the internet, spaceranger uses the value of the --slide argument to automatically download the slide layout file in gpr format.
spaceranger count --id="Visium_FFPE_Mouse_Brain" \
                  --description="Adult Mouse Brain (FFPE) using Mouse WTA Probe Set" \
                  --transcriptome=refdata-gex-mm10-2020-A \
                  --probe-set=Visium_FFPE_Mouse_Brain_probe_set.csv \
                  --fastqs=datasets/Visium_FFPE_Mouse_Brain_fastqs \
                  --image=datasets/Visium_FFPE_Mouse_Brain_image.jpg \
                  --slide=V11J26-127 \
                  --slidefile=V11J26-127.gpr \
                  --area=B1 \
                  --reorient-images=true \
                  --localcores=16 \
                  --localmem=128
In absence of internet connectivity to the compute platform, you can download this specific slide layout file in gpr format and provide it to spaceranger using the --slidefile argument.

Below are brief descriptions of the above command line options.

Option Description
--id The id must be unique string and will be used to name the resulting folder with all of the pipeline outputs. We choose to keep the original dataset name of Visium_FFPE_Mouse_Brain
--description This is sample description included in the output files (e.g. web_summary.html). We describe the sample as "Adult Mouse Brain (FFPE) using Mouse WTA Probe Set"
--transcriptome The path to the species specific pre-compiled transcriptome files. Note that you can either provide the relative path as shown above or the absolute path to this folder. As the tissue sample was of mouse origin, we provide the path to the mouse reference transcriptome refdata-gex-mm10-2020-A
--probe-set The absolute or relative path to the species specific probe set reference file in CSV format. Since the tissue sample is derived from mouse, we specify the relative path to the mouse probe set reference as Visium_FFPE_Mouse_Brain_probe_set.csv
--fastqs The path to the folder containing sample sequencing files in FASTQ format. The path can be relative as shown above or absolute. The relative path is /datasets/Visium_FFPE_Mouse_Brain_fastqs
--image The path to a single brightfield image with H&E staining in either TIFF or JPEG formats. The path can be relative or absolute. Here we have a TIFF format image with the following relative path /datasets/Visium_FFPE_Mouse_Brain_image.jpg
--slide The visium slide serial number of which the tissue sample was mounted and the value here is V11J26-127
--area The capture area identifier on the visium slide. It can be one of four values: A1, B1, C1 or D1. Here the tissue sample was mounted on B1 capture area.
--slidefile The slide layout file in gpr format which is provided when spaceranger does not have internet access. You can download the slide layout file and provide it as V11J26-127.gpr
--reorient-images Option to choose whether spaceranger should rotate and mirror the image to find the best fiducial alignment. Acceptable values are true or false with default being true. Useful to set to false when you are absolutely certain the fiducial corners in the input image are in the canonical positions (hourglass in top left corner) similar to included image Visium_FFPE_Mouse_Brain_image.jpg. Setting this to false will reduce pipeline runtime and prevent the pipeline from finding a fiducial alignment where the image is rotated/mirrored.
--localcores The number of CPU cores available to run the spaceranger count pipeline. The maximum upper limit for your specific compute system is determined using the sitecheck subcommand. We will use 16 cores in this tutorial.
--localmem The max memory in GB available to run the spaceranger count pipeline. The maximum upper limit for your specific compute system is determined using the sitecheck subcommand. We will use 128 GB in this tutorial.

Run spaceranger count

At the start of the pipeline, you should see the message about the preflight checks printed to the command line.

# Run spaceranger count
spaceranger count --id="Visium_FFPE_Mouse_Brain" \
                  --description="Adult Mouse Brain (FFPE) using Mouse WTA Probe Set" \
                  --transcriptome=refdata-gex-mm10-2020-A \
                  --probe-set=Visium_FFPE_Mouse_Brain_probe_set.csv \
                  --fastqs=datasets/Visium_FFPE_Mouse_Brain_fastqs \
                  --image=datasets/Visium_FFPE_Mouse_Brain_image.jpg \
                  --slide=V11J26-127 \
                  --area=B1 \
                  --reorient-images=true \
                  --localcores=16 \
                  --localmem=128
# Run spaceranger count
spaceranger count --id="Visium_FFPE_Mouse_Brain" \
                  --description="Adult Mouse Brain (FFPE) using Mouse WTA Probe Set" \
                  --transcriptome=refdata-gex-mm10-2020-A \
                  --probe-set=Visium_FFPE_Mouse_Brain_probe_set.csv \
                  --fastqs=datasets/Visium_FFPE_Mouse_Brain_fastqs \
                  --image=datasets/Visium_FFPE_Mouse_Brain_image.jpg \
                  --slide=V11J26-127 \
                  --slidefile=V11J26-127.gpr \
                  --area=B1 \
                  --reorient-images=true \
                  --localcores=16 \
                  --localmem=128
Martian Runtime - v4.0.5
 
Running preflight checks (please wait)...
Checking sample info...
Checking FASTQ folder...
Checking reference...
Checking reference_path...
Checking optional arguments...
 
...

Successful completion of the pipeline is indicated by summary of the output files generated.

Outputs:
- Run summary HTML:                                          /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/web_summary.html
- Outputs of spatial pipeline:
    aligned_fiducials:              /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/spatial/aligned_fiducials.jpg
    detected_tissue_image:          /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/spatial/detected_tissue_image.jpg
    scalefactors_json:              /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/spatial/scalefactors_json.json
    tissue_hires_image:             /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/spatial/tissue_hires_image.png
    tissue_lowres_image:            /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/spatial/tissue_lowres_image.png
    cytassist_image:                null
    aligned_tissue_image:           null
    tissue_positions:               /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/spatial/tissue_positions.csv
    spatial_enrichment:             /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/spatial/spatial_enrichment.csv
    barcode_fluorescence_intensity: null
- Run summary CSV:                                           /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/metrics_summary.csv
- Correlation values between isotypes and Antibody features: null
- BAM:                                                       /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/possorted_genome_bam.bam
- BAM BAI index:                                             /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/possorted_genome_bam.bam.bai
- BAM CSI index:                                             null
- Filtered feature-barcode matrices MEX:                     /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/filtered_feature_bc_matrix
- Filtered feature-barcode matrices HDF5:                    /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/filtered_feature_bc_matrix.h5
- Unfiltered feature-barcode matrices MEX:                   /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/raw_feature_bc_matrix
- Unfiltered feature-barcode matrices HDF5:                  /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/raw_feature_bc_matrix.h5
- Secondary analysis output CSV:                             /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/analysis
- Per-molecule read information:                             /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/molecule_info.h5
- Loupe Browser file:                                        /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/cloupe.cloupe
- Feature Reference:                                         null
- Target Panel file:                                         null
- Probe Set file:                                            /spaceranger_tutorial/Visium_FFPE_Mouse_Brain/outs/probe_set.csv
 
Waiting 6 seconds for UI to do final refresh.
Pipestance completed successfully!

Explore the output of spaceranger count

After the run is completed, the working directory will have a new folder named Visium_FFPE_Mouse_Brain (value provided to --id argument) that contains all the metadata and outputs generated from the spaceranger count pipeline. We will highlight some key components of this folder (highlighted in bold):

Visium_FFPE_Mouse_Brain
├── _cmdline
├── _filelist
├── _finalstate
├── _invocation
├── _jobmode
├── _log
├── _mrosource
├── outs
├── _perf
├── _sitecheck
├── SPATIAL_RNA_COUNTER_CS
├── _tags
├── _timestamp
├── _uuid
├── Visium_FFPE_Mouse_Brain.mri.tgz
├── _vdrkill
└── _versions

The outs folder contain all the calculated results.


You can further explore and understand these results by

FAQs

Q. Q. I ran spaceranger count and got this error Could not retrieve spot layout data. What does this mean and how can I proceed ?

When you specify the visium slide id using the --slide argument, spaceranger count downloads the corresponding slide file layout file in gpr format. This step requires internet connectivity. However in some instances, compute platforms may not have internet access and hence the resulting error message. If you know the visium slide id, you can download the slide layout file and provide it to the pipeline using the --slidefile argument along with specifying the capture area with --area.