Single Cell CNV
Cell Ranger DNA, printed on 10/19/2020
Release Notes for 1.1
Cell Ranger DNA 1.1.0 (2019-05-14)
- Include new cellranger-dna bamslice pipeline to allow subsetting BAMs using a specified set of cellular barcodes.
- Include new cellranger-dna reanalyze pipeline to allow reprocessing data from a finished pipeline with optional parameters, subsets of data, or user-defined structure inputs.
- Include new cellranger-dna aggr pipeline to allow aggregating data from multiple pipelines and analyzing it as one dataset.
- Refactor read processing stages to greatly increase speed and reduce I/O load.
- Improve event contiguity by reducing the mappability floor from 90% to 70%, which increases the fraction of genome mappable by 1-3%.
- Include a mappability term in the models to compensate for the greater range of mappability, and to generally improve event sensitivity and contiguity.
- Determine the ploidy scaling of a group of cells solely from the ploidy scalings of its constituent cells. This prevents unintended scaling abnormalities that were occasionally observed with Cell Ranger DNA version 1.0.
- Reformulate the CNV segment confidence score to improve sensitivity to variance and bias.
- Update websummary, including reformatting graphs and metrics to facilitate easier information transfer.
- In the hero metrics, replace "Median effective reads per Mb" with "Median estimated CNV resolution (in Mb)". This metric indicates the approximate size of copy number events (in Mb) that can be profiled accurately in single cells based on the sequencing depth and library quality, as determined by a theoretical model. In accordance with this change, remove the warning when "Median effective reads per Mb" < 50.
- Report R1 Length and R2 Length, which capture the median length in base pairs of read1 and read2, respectively.
- Include additional metadata in the
cnv_data.h5 including the pipeline run, reference chosen, and
reanalyze configuration if applicable.
DC tag from the output BAM file.