10x Genomics
Chromium Single Cell Gene Expression

Cell Ranger4.0, printed on 11/21/2024

Demultiplexing FASTQs with bcl2fastq

cellranger mkfastq is the preferred option for converting BCLs to Cell Ranger FASTQs. You can also use Illumina bcl2fastq directly to generate FASTQs. Choose this method if bcl2fastq is tightly integrated into your sequencing workflow or if you want more control over demultiplexing parameters.

Demultiplexing Chromium data with Illumina bcl2fastq requires the correct specification of the sample sheet and command-line options. This guide will walk you through what you'll need to do to generate Cell Ranger-compatible FASTQs.

Dual Indexing

This section describes how to configure bcl2fastq for libraries created with the Dual Index Plate TT, Set A or Dual Index Plate NT, Set A.

These are 'unique dual-indexing' sample indexes. This means that there is a unique sample index barcode in the both the i7 and i5 index reads (also known as I1 and I2). When demultiplexing flowcells where both index reads have been sequenced, bcl2fastq requires that both index sequences match the expected sequence for a read to be assigned to that sample. This solves the 'index hopping' issue present on Illumina patterned flowcell sequencers.

Sample Sheet Creation

You can download the Sample Index Reference files for dual indexing kits here:

• Dual Index Plate TT, Set A (Gene Expression): CSV | JSON
• Dual Index Plate NT, Set A (Feature Barcode): CSV | JSON

The index sequence in the sample index reference file should be entered into the index column of the bcl2fastq sample sheet.

Either the index2_workflow_a or index2_workflow_b sequence should be entered into the index2 column of the bcl2fastq sample sheet, depending on the sequencing instrument in use.

• index2_workflow_a: NovaSeq™ 6000, MiSeq™, HiSeq 2500, and HiSeq 2000.
• index2_workflow_b: iSeq™ 100, MiniSeq™, NextSeq™, HiSeq X, HiSeq 3000/4000,

More information about dual-indexing is available in the Illumina Indexed Sequencing Overview Guide

The Illumina Experiment Manager can also be used to create sample sheets for use with bcl2fastq.

Example Sample Sheet

When you plan an experiment, you should know the name of the sample index set used for each sample, which comes from the reagent kit (such as "SI-TT-A2"). For each sample, enter its lane, sample name, and sample index set into the Illumina bcl2fastq sample sheet. Here is a bcl2fastq sample sheet for a HiSeq 2500:

[Header]
EMFileVersion,4

[Reads]
28
90

[Data]
Lane,Sample_ID,Sample_Name,index,index2,Sample_Project,Original_Sample_ID
1,test_sample,test_sample,TGGTCCCAAG,CCTCTGGCGT,H5T2YBCX3,test_sample

Single Indexing

This section describes how to configure bcl2fastq for libraries created with the Single Index Kit T Set A.

Sample Sheet Generator

You will need to create a sample sheet in order to get bcl2fastq to correctly embed the names of samples into output FASTQ files. There is a key difference to keep in mind when creating sample sheets for a Chromium run. Each Chromium sample index set is actually a blend of 4 different sequence oligos, and each oligo must be represented as a separate row in the sample sheet. This means that for every sample being demultiplexed from the flowcell, there should be 4 lines in the sample sheet.

The tool below will help you accurately generate data lines for your sample sheet. When you plan an experiment, you should know the name of the sample index set used for each sample, which comes from the reagent kit (such as "SI-P01-A2"). For each sample, enter its lane, sample name, and sample index set below, and then press 'Add'. When you're done, you can either copy and paste comma-separated output directly into a text editor to create a sample sheet CSV, or copy/paste tab-separated output into a spreadsheet such as Microsoft Excel.

Please note that the default R1/R2 lengths differ between the v2 and v3 chemistries:

If you are just running a single sample in a lane, then you can have a single line with the index blank, though bcl2fastq will include reads associated with any sample index.

Running bcl2fastq

Illumina bcl2fastq must be called with the correct --use-bases-mask argument and other arguments in order to properly demultiplex and output FASTQs for all the reads in a Chromium library. In the examples below, ${FLOWCELL_DIR} is the directory that contains a flowcell's Data folder, ${OUTPUT_DIR} is the directory that you want to output FASTQs to, and ${SAMPLE_SHEET_PATH} is the path to the sample sheet CSV you created.

bcl2fastq Version 2.17 or higher

This is the most common case, for sequencers running RTA 1.18.54 and higher.

$ bcl2fastq --use-bases-mask=Y26,I8,Y98 \
  --create-fastq-for-index-reads \
  --minimum-trimmed-read-length=8 \
  --mask-short-adapter-reads=8 \
  --ignore-missing-positions \
  --ignore-missing-controls \
  --ignore-missing-filter \
  --ignore-missing-bcls \
  -r 6 -w 6 \
  -R ${FLOWCELL_DIR} \
  --output-dir=${OUTPUT_DIR} \
  --interop-dir=${INTEROP_DIR} \
  --sample-sheet=${SAMPLE_SHEET_PATH}

Version 1.8.4

$ configureBclToFastq.pl --use-bases-mask=Y26,I8,Y98 \
  --fastq-cluster-count=20000000 \
  --no-eamss \
  --ignore-missing-bcl \
  --ignore-missing-control \
  --ignore-missing-stats \
  --mismatches=1 \
  --input-dir=${FLOWCELL_DIR}/Data/Intensities/BaseCalls \
  --output-dir=${OUTPUT_DIR}

In both cases, if you want to limit bcl2fastq to a subset of lanes, you will need to supply values to the --tiles argument.

Omitting Extra Bases from Reads

If you add extra bases to a sample index read, you will need to account for this in the --use-bases-mask argument. For example, if you ran a sample index read with 9 bases, you will need to truncate the last base in order for Cell Ranger to run correctly.

You can exclude a single base by adding a single n character to the read argument, or adding n* to exclude all bases after a certain position. See below:

Running Cell Ranger with bcl2fastq FASTQs

After generating FASTQs, you should be able to follow the pipeline instructions, with one caveat. Instead of using the --indices argument to cellranger to select samples, you will use the --sample argument. The value of --sample should be the name of the sample, which should have been in the Sample_Name column in your sample sheet. The value of --fastqs should be ${OUTPUT_DIR}/${PROJECT_NAME} where ${OUTPUT_DIR} is as defined above and ${PROJECT_NAME} is the value in the Sample_Project column in the sample sheet.