Question: What is the difference between mkfastq, bcl2fastq, and demux? Which method should I use to demultiplex my 10x libraries?
Answer: Raw sequence data (BCL files) from all 10x pipelines can be demultiplexed the same way. All 10x Genomics pipelines come bundled with a
mkfastq pipeline (eg
supernova mkfastq, etc). This is a thin wrapper around Illumina's
bcl2fastq which does the actual demultiplexing. The main advantage of using the mkfastq pipeline is the automatic recognition of 10x index sets via a simple CSV sample sheet with only three columns (Lane, Sample, Index). Using the simple CSV sample sheet format allows you to avoid the more complex and syntax-error prone IEM format that
bcl2fastq normally requires. You can also pass through options and arguments specific to
bcl2fastq. For more information about
bcl2fastq, please see Illumina's page.
mkfastq is optional - you can also use
bcl2fastq directly if you prefer, or you may have received FASTQ files from a sequencing core or collaborator that were demultiplexed directly by
For more information about the
mkfastq pipeline please see the link for each respective product:
mkfastq we had the
demux pipeline, which is now deprecated, but many datasets that were demultiplexed with this pipeline are still publicly available. For more information on
demux, please see this page.