Downsampling

DRAGEN RNA also supports internal downsampling, which is a process by which a random sub-sample of reads is selected from the dataset after trimming and alignment for downstream analysis. In RNA-Seq, this can be useful in two ways - it can speed up the analysis of samples with excessively high coverage, and it can allow for more accurate cross-comparisons between different samples.

If --enable-down-sampler is set to true and a value specified for --down-sampler-reads, DRAGEN will use only that many RNA fragments (including both Read 1 and Read 2) for quantification, fusion, variant calling and output to BAM. The entire input dataset is still used for the generation of trimming, fastqc, and mapping metrics.