Support homeCell RangerTutorials
Running Cell Ranger multi

Running Cell Ranger multi

This tutorial is written with Cell Ranger v7.0.0. Starting with Cell Ranger v8.0, it is mandatory to use the --create-bam parameter when executing the cellranger count and cellranger multi pipelines. This new parameter replaces the previously used --no-bam option. All other arguments remain compatible with newer versions, unless otherwise specified.

This tutorial describes how to run the cellranger multi pipeline (we recommend completing the other Cell Ranger pipeline tutorials in this series first).

The example data used in this tutorial is for a 3' Cell Multiplexing dataset. In Cell Ranger v7.0 and later, Single Cell Fixed RNA Profiling datasets can be analyzed with the cellranger multi pipeline as well. For specific multi pipeline details and outputs, see:

  • Running multi for Fixed RNA Profiling
  • Fixed RNA Profiling outputs

In this tutorial, we will analyze a 3' Cell Multiplexing dataset that consists of two cell lines, Jurkat and Raji, multiplexed at equal proportions with one CMO per cell line, resulting in a pooled sample labeled with two CMOs. Gene Expression (GEX) and Cell Multiplexing libraries were prepared with the Chromium Next GEM Single Cell 3ʹ Reagent Kits v3.1 (Dual Index) with Feature Barcode technology.

Use wget to download the FASTQ data (about 44 GB):

wget https://cg.10xgenomics.com/samples/cell-exp/6.0.0/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_Multiplex/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_Multiplex_fastqs.tar

As of this tutorial's publication, the most current reference transcriptome was the Human reference (GRCh38) - 2020-A. Download and decompress it:

wget https://cf.10xgenomics.com/supp/cell-exp/refdata-gex-GRCh38-2020-A.tar.gz tar -xf refdata-gex-GRCh38-2020-A.tar.gz

Decompress the FASTQ files:

tar -xf SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_Multiplex_fastqs.tar

Navigate to the FASTQ files and observe their filenames. There is one directory that contains the FASTQ files for the GEX library. There are two that contain FASTQ files for the Cell Multiplexing Capture library because the same physical library was sequenced twice for this particular dataset - first for a preliminary sample quality check and second for the actual analysis. The simplest scenario is to analyze one Gene Expression and one Multiplexing Capture library, which we will demonstrate using the FASTQ files in the ..._1_gex and ..._1_multiplexing_capture directories.

. ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L001_I1_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L001_I2_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L001_R1_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L001_R2_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L002_I1_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L002_I2_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L002_R1_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L002_R2_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L003_I1_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L003_I2_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L003_R1_001.fastq.gz │ └── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex_S2_L003_R2_001.fastq.gz ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L001_I1_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L001_I2_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L001_R1_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L001_R2_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L002_I1_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L002_I2_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L002_R1_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L002_R2_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L003_I1_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L003_I2_001.fastq.gz │ ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L003_R1_001.fastq.gz │ └── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture_S1_L003_R2_001.fastq.gz └── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_2_multiplexing_capture ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_2_multiplexing_capture_S1_L001_I1_001.fastq.gz ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_2_multiplexing_capture_S1_L001_I2_001.fastq.gz ├── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_2_multiplexing_capture_S1_L001_R1_001.fastq.gz └── SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_2_multiplexing_capture_S1_L001_R2_001.fastq.gz

The cellranger multi pipeline has two inputs:

  • --id is used to name the output directory that the pipeline runs in.
  • --csv takes a CSV file that points to the FASTQ files, and contains other parameters from the cellranger multi pipeline.

In this tutorial, you only need to edit a few lines in a pre-made CSV using a text editor of your choice. We are using the text editor nano to edit the CSV:

nano SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K.csv

Copy and paste the code block below into your text editor. For this example, there are three sections: [gene-expression], [libraries], and [samples]. Important: replace the /path/to/ text with the full paths to the reference (in [gene-expression] section) and FASTQ files (in [libraries] section) that you downloaded before saving the CSV file.

[gene-expression] ref,/path/to/refdata-gex-GRCh38-2020-A create-bam,true [libraries] fastq_id,fastqs,lanes,feature_types SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex,/path/to/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex,any,Gene Expression SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture,/path/to/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture,any,Multiplexing Capture [samples] sample_id,cmo_ids,description Jurkat,CMO301,Jurkat Raji,CMO302,Raji

Alternatively, download a multi config CSV template and customize it. Cell Ranger v7.1 enables users to download a multi config CSV template by running:

cellranger multi-template --output=/path/to/FILE.csv

Replace path above with the path to the directory in which you wish to output the template. Omitting the file path downloads the file into your working directory. After downloading, customize the template as shown above.

To print a list and description of all configurable parameters available in cellranger multi, run:

cellranger multi-template --parameters

Learn more about the multi config CSV on the running cellranger multi page, which also describes all the available sections, fields, and optional parameters.

Optional: If you want to analyze both sequence runs of the Multiplexing Capture library from this example dataset, add an additional line to the [libraries] section for the 2nd sequence run:

[libraries] fastq_id,fastqs,lanes,feature_types SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex,/path/to/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K/ SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_gex,any,Gene Expression SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture,/path/to/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_1_multiplexing_capture,any,Multiplexing Capture SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_2_multiplexing_capture,/path/to/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K/SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K_2_multiplexing_capture,any,Multiplexing Capture

Next run the cellranger multi command with --help to get the usage and a full list of modifiable parameters.

cellranger multi --help

The output looks similar to this:

cellranger-multi Analyze multiplexed data or combined gene expression/immune profiling/feature barcode data USAGE: cellranger multi [OPTIONS] --id <ID> --csv <CSV> OPTIONS: --id A unique run id and output folder name [a-zA-Z0-9_-]+ --description &ltTEXT&gt Sample description to embed in output files [default: ] --csv &ltCSV&gt Path of CSV file enumerating input libraries and analysis parameters --dry Do not execute the pipeline. Generate a pipeline invocation (.mro) file and stop ...

To run cellranger multi, enter a command such as:

cellranger multi --id=Jurkat_Raji_10K --csv=SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K.csv

Cell Ranger 6.0+ should start with a message like this:

Martian Runtime - v4.0.8 Running preflight checks (please wait)...

Depending on your computational resources, it may take some time for the pipeline to complete. When it does, it should conclude with a message like this:

Waiting 6 seconds for UI to do final refresh. Pipestance completed successfully! 2022-05-09 19:20:44 Shutting down.

Next, examine the output files using the tree command:

cd Jurkat_Raji_10K/outs tree

The tree command will list 73 directories with 96 files. (If you are used to a cellranger count run, recall that multiplexing two samples necessitates doubling the per-sample outputs, and these numbers will grow correspondingly as more samples are multiplexed into a single GEM well). Additionally, some output files are general to the entire experiment rather than a specific CMO.

The first section of the outputs contains the config.csv file, a duplicate of the input config CSV (SC3_v3_NextGem_DI_CellPlex_Jurkat_Raji_10K.csv). The multi directory contains a count directory and a multiplexing_analysis directory:

└─ config.csv └─ multi ├── count │ ├── feature_reference.csv │ ├── raw_cloupe.cloupe │ ├── raw_feature_bc_matrix │ │ ├── barcodes.tsv.gz │ │ ├── features.tsv.gz │ │ └── matrix.mtx.gz │ ├── raw_feature_bc_matrix.h5 │ ├── raw_molecule_info.h5 │ ├── unassigned_alignments.bam │ └── unassigned_alignments.bam.bai └── multiplexing_analysis ├── assignment_confidence_table.csv ├── cells_per_tag.json ├── tag_calls_per_cell.csv └── tag_calls_summary.csv

For more information on these files, see Cell Multiplexing Outputs.

The per_sample_outs directory contains two directories, one for Jurkat and one for Raji. For brevity, only the Jurkat outputs are shown here.

In the Jurkat/count/analysis directory, the clustering directory contains CSV files with the results of graph-based clusters and K-means clustering from 2-10:

└── clustering ├── graphclust │ └── clusters.csv ├── kmeans_10_clusters │ └── clusters.csv ├── kmeans_2_clusters │ └── clusters.csv ├── kmeans_3_clusters │ └── clusters.csv ├── kmeans_4_clusters │ └── clusters.csv ├── kmeans_5_clusters │ └── clusters.csv ├── kmeans_6_clusters │ └── clusters.csv ├── kmeans_7_clusters │ └── clusters.csv ├── kmeans_8_clusters │ └── clusters.csv └── kmeans_9_clusters └── clusters.csv

The diffexp directory likewise contains CSV files with the results of differential expression analysis between the clusters reported above:

└── diffexp ├── graphclust │ └── differential_expression.csv ├── kmeans_10_clusters │ └── differential_expression.csv ├── kmeans_2_clusters │ └── differential_expression.csv ├── kmeans_3_clusters │ └── differential_expression.csv ├── kmeans_4_clusters │ └── differential_expression.csv ├── kmeans_5_clusters │ └── differential_expression.csv ├── kmeans_6_clusters │ └── differential_expression.csv ├── kmeans_7_clusters │ └── differential_expression.csv ├── kmeans_8_clusters │ └── differential_expression.csv └── kmeans_9_clusters └── differential_expression.csv

The pca, tsne, and, umap directories contain CSV files for dimensionality reduction:

├── pca │ └── 10_components │ ├── components.csv │ ├── dispersion.csv │ ├── features_selected.csv │ ├── projection.csv │ └── variance.csv ├── tsne │ ├── 2_components │ │ └── projection.csv │ └── multiplexing_capture_2_components │ └── projection.csv └── umap ├── 2_components │ └── projection.csv └── multiplexing_capture_2_components └── projection.csv

The remaining per_sample_outs are described in the Cell Multiplexing Outputs.

├── sample_cloupe.cloupe ├── feature_reference.csv ├── sample_alignments.bam ├── sample_alignments.bam.bai ├── sample_filtered_barcodes.csv ├── sample_filtered_feature_bc_matrix │ ├── barcodes.tsv.gz │ ├── features.tsv.gz │ └── matrix.mtx.gz ├── sample_filtered_feature_bc_matrix.h5 ├── sample_molecule_info.h5 ├── metrics_summary.csv └── web_summary.html

Questions or feedback about this tutorial? Contact [email protected].