New features

• Cell Ranger ARC v2.1 introduces the ability to perform automated cell type annotations using Gene Expression libraries (not ATAC). This beta feature is available as part of the cellranger-arc count pipeline.
  • A 10x Genomics Cloud Analysis account is required to perform automated cell type annotations, as the process is conducted within Cloud Analysis.
  • Users must agree to the 10x Genomics End User License Agreement (EULA) before using this feature.
  • Use the cellranger-arc cloud auth setup tool to simplify the retrieval of the cloud account security token, a required input for performing automated cell type annotations.
  • Learn more about running cell annotation and interpreting annotation output files.
• Cell Ranger ARC v2.1 includes anonymized telemetry data collection to help 10x Genomics improve product functionality by providing insights into usage patterns, diagnosis, and potential feature improvements. For more information on the collected data and usage, visit the Cell Ranger ARC Telemetry page.
• Added a batch correction algorithm to cellranger-arc aggr. See the aggr and algorithms pages for more details.

Bug fixes & general improvements

• Users can expect to see substantial performance improvements, particularly runtime for cellranger-arc aggr runs. See the system requirements page for updated performance information.
• Users can now invoke the --nosecondary parameter to disable secondary analysis and speed up runtime. See the command line arguments page for more information.
• The --create-bam argument is now required for cellranger-arc count.
• To enable analysis of large datasets, the 20,000 total cell limit has been removed. (Previously, there was a hard-coded limit of 20,000 cells for any cellranger-arc analysis).
• In all web_summary.html files, only UMAP projections will be shown. In the cellranger-arc aggr pipeline, t-SNE has been replaced with UMAP.
• Strand information is now included in the fragments.tsv file, determined from R1/R2 adapter sequences. This enhancement improves the performance of downstream filters, specifically doublet and inclusion list contamination detection.
• Cell Ranger ARC v2.1 can now ingest FASTQs with a quality score up to the full supported range (93 instead of 41).

New system requirements

• AVX instructions must be available on CPUs which are running Cell Ranger ARC v2.1 or higher. AVX instructions were introduced in 2011 on both Intel and AMD CPUs. In the future, AVX2 will be required. AVX2 instructions were introduced in 2013 on Intel CPUs and in 2015 on AMD CPUs.

Deprecated & unsupported features

• Ubuntu 14.04 and Westmere CPUs are no longer supported.
• The cellranger-arc mkfastq pipeline is deprecated. Please demultiplex directly with Illumina software.

• 2024-A reference transcriptomes for human (GRCh38) and mouse (GRCm39) are now available for download. See Reference Release Notes.

• Bug fixes
  • Fixes a bug where an upgrade to Illumina NovaSeq control software v1.8 (reagent name change in recipe XML file) resulted in a silent cellranger-arc mkfastq error and a significant number of reads going into Undetermined/ because the orientation of i5 (Index2) could not be autodetected

• Bug fixes
  • Fixes an issue where the metric "post-normalization number of reads" in the aggr web summary was incorrect.
  • Fixes an issue where integer overflow could occur on large aggr or reanalyze runs.
  • Fixes an issue where sample indices in the FASTQ header line were ignored unless the indices were reverse complemented.
  • Fixes an issue where gex_umis_count in per_barcode_metrics.csv were all zeros.
  • The --nthreads parameter has been re-enabled for cellranger-arc mkref.

• New wavelet-based peak caller:

  • Eliminates large peaks much larger than 5kb. Peaks have a tighter size distribution around ~ 1kb.
  • Improves detection of cluster-specific peaks.
  • Improves reproducibility between technical replicates.
  • Consistent performance across a range of cell loads.
  • Fixes crashes in the signal-background fitting procedure.

• Changed the duplicate marking algorithm:

  • Two read pairs are duplicates if they share the same start, end and cell barcode. Previously, only the start and end were used.
  • Boosts median fragments per cell by as much as 25% at high cell loads.
  • PCR and sequencer duplicates are no longer distinguished.

• Improved computational performance:

  • Up to 4x faster, 0.5x disk requirements.
  • Complete rewrite of read processing and differential accessibility analysis in Rust.
  • Minimize disk I/O.

• Added the aggr pipeline to aggregate multiple libraries from multiple GEM wells.

• Added the reanalyze pipeline to enable customized reanalysis.

• Changed cell caller override: when --min-atac-count=X and --min-gex-count=Y is specified we choose all barcodes with atac count >= X and (previously or) gex count >= Y.

• The raw feature barcode matrix now only contains barcodes with at least one ATAC count or one GEX count; previously, all allowed barcodes were present.

• Change in barcode error correction: correct Hamming distance 1 errors (previously 2).

• Added header lines beginning with # to the fragments.tsv.gz and peaks.bed files that contain version, reference and sample information.

• Change to ATAC peak annotation TSV format:

  • The peak column is now split into three columns: chrom, start, end.
  • When a peak has multiple gene annotations, the same peak appears in multiple rows with each annotation. Previously, each row represented one peak and multiple annotations were expressed using ; separators in the same row.

• Improvements to reference construction:

  • Restrictions on the number of contigs or gene-containing contigs (primary contigs) in the reference have been eliminated.
  • Bug fix for GTFs that do not contain a gene_name attribute.
  • Contig names are allowed to contain - or _ characters.
  • Eliminated discrepancies between reference checks in mkref and preflight checks in count. Previously, it was possible to pass checks in mkref and fail checks in count.

• Eliminated secondary alignments from the position-sorted BAM.

• Loupe browser files generated by the pipeline can only be opened by Loupe browser version 5.0 or later.

• Increases the upper limit on primary contigs (those that have at least one gene annotation) from 100 to 500. The pipeline will error out if more than 1000 total contigs are present in the reference.
• Disables multithreading in mkref to address an issue where mkref would fail on hardware without AVX support. This will be fixed in a future release.

• Creates cellranger-arc count for the analysis of 10x Chromium Epi Multiome ATAC + Gene Expression data generated from a single GEM well.
• Creates cellranger-arc mkfastq for demultiplexing of ATAC or GEX flow cell data for analysis.
• Creates cellranger-arc mkref to construct reference packages for use with cellranger-arc count starting with a reference FASTA file and a gene annotations GTF file.
• Note: the software cannot be used for the analysis of 10x Chromium Single Cell ATAC, 10x Chromium Single Cell 3' Gene Expression data, or for any kind of joint analysis of the two.