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

10x Genomics
Chromium Single Cell Immune Profiling

Filtered Outputs

The per_samples_outs/ directory is produced after a successful execution of the multi pipeline and contains filtered data, i.e., data from cell-associated barcodes in this sample. These are the main outputs of interest.

Contents of the following folders located within the per_samples_outs/ directory are described here. Click on the folder name below or scroll down to learn more.

Refer to the count and vdj pages for detailed explanations.


The count/ folder contains the results of 5' single cell gene expression analysis. The count directory looks like this:

├── count
    ├── analysis
    ├── cloupe.cloupe
    ├── sample_alignments.bam
    ├── sample_alignments.bam.bai
    ├── sample_barcodes.csv
    ├── sample_feature_bc_matrix
    ├── sample_feature_bc_matrix.h5
    └── sample_molecule_info.h5
File/Folder Description
analysis Folder containing the results of graph-based clusters and K-means clustering 2-10; differential gene expression analysis between clusters; and PCA, t-SNE, and UMAP dimensionality reduction. Learn more
cloupe.cloupe A Loupe Browser readable file.
sample_alignments.bam Indexed BAM file containing position-sorted reads aligned to the genome and transcriptome, as well as unaligned reads. Learn more
sample_alignments.bam.bai Companion file to the sample_alignment.bam that serves as an external index
sample_barcodes.csv File containing a list of barcodes associated with aligned reads. The barcode sequence ends in a suffix with a dash separator followed by a number. The number denotes a GEM well, and is used to virtualize barcodes in order to achieve a higher effective barcode diversity when combining samples generated from separate GEM chip channel runs. The number should be “1” across all barcodes when analyzing a sample from a single GEM well. The suffix-based preservation of GEM well information is especially useful when running cellranger aggr on multiple libraries generated from different GEM chip channels.
sample_feature_bc_matrix Contains only detected cell-associated barcodes. Each element of the matrix is the number of UMIs associated with a feature (row) and a barcode (column). This file can be input into third-party packages and allows users to wrangle the barcode-feature matrix (e.g. to filter outlier cells, run dimensionality reduction, normalize gene expression). This file is similar to the filtered_feature_bc_matrix file described here
sample_feature_bc_matrix.h5 Same information as sample_molecule_bc_matrix in H5 format.
sample_molecule_info.h5 Contains per-molecule information for all molecules that contain a valid barcode and valid UMI and were assigned with high confidence to a gene or Feature Barcode. This file is a required input to run cellranger aggr. Learn more

vdj (_t/_b)

The vdj_t/ and vdj_b/ folders contain the results of V(D)J immune profiling analysis for T cells and B cells, respectively. The output file names and file structure in the vdj_b/ and vdj_t/ folders are identical, and are only described once. The vdj_t and vdj_b directories have this structure:

in this list we are missing all_contig_annotations.bed/csv/json/ files; all_contig.bam/.bam.bai/fasta/.fasta.fai/.fastq files. Is that by design?
├── vdj_b/t
    ├── airr_rearrangement.tsv
    ├── cell_barcodes.json
    ├── clonotypes.csv
    ├── concat_ref.bam
    ├── concat_ref.bam.bai
    ├── concat_ref.fasta
    ├── concat_ref.fasta.fai
    ├── consensus_annotations.csv
    ├── consensus.bam
    ├── consensus.bam.bai
    ├── consensus.fasta
    ├── consensus.fasta.fai
    ├── filtered_contig_annotations.csv
    ├── filtered_contig.fasta
    ├── filtered_contig.fastq
    ├── vdj_contig_info.pb
    └── vloupe.vloupe
File/Folder Description
airr_rearrangement.tsv Annotated contigs and consensus sequences of V(D)J rearrangements in the AIRR format. Learn more
cell_barcodes.json List of barcodes identified as T/B cells.
clonotypes.csv High-level descriptions of each clonotype. Learn more
concat_ref.bam For each clonotype consensus, each reference sequence is the annotated germline segments concatenated together. This file shows how both the per-cell contigs and the clonotype consensus contig relate to the germline reference. concat_ref.bam is expected to reveal polymorphisms, somatic mutations, and recombination-induced differences such as non-templated nucleotide additions.
concat_ref.bam.bai Companion file to the concat_ref.bam that serves as an external index.
concat_ref.fasta Concatenated V(D)J reference segments for the segments detected on each consensus sequence. These serve as an approximate reference for each consensus sequence.
concat_ref.fasta.fai Companion file to the concat_ref.fasta that serves as an external index.
consensus_annotations.csv High-level and detailed annotations of each clonotype consensus sequence.
consensus.bam Each reference sequence is a clonotype consensus sequence, and each record is an alignment of a single cell's contig against this consensus. For a clonotype consensus sequence, this file shows how the constituent per-cell assemblies support the consensus.
consensus.bam.bai Companion file to the consensus.bam that serves as an external index.
consensus.fasta Clonotype consensus sequences.
consensus.fasta.fai Companion file to the consensus.fasta that serves as an external index.
filtered_contig_annotations.csv High-level annotations of each high-confidence, cellular contig. This is a subset of all_contig_annotations.csv. Learn more
filtered_contig.fasta High-confidence contig sequences in cell barcodes in FASTA format.
filtered_contig.fastq High-confidence contig sequences in cell barcodes in FASTQ format.
vdj_contig_info.pb This file stores the contig annotations, V(D)J reference and additional metadata in a protobuf binary file format. This file is required to run the cellranger aggr pipeline. Learn more
vloupe.vloupe Loupe V(D)J Browser readable file.

Next Steps