Workflows

bcl2fastq

Inputs

• bcl: gs:// path to BCL
• samplesheet: gs:// path to samplesheet
• technique: "bcl2fastq" or "cellranger" or "cellranger-arc" or "cellranger-atac"
• fastq_output_path (optional): gs:// path to write fastqs (default fastqs/{basename(bcl)})
• log_output_path (optional): gs:// path to write logs (default logs/{basename(bcl)})

Commands

• cellranger mkfastq --run=BCL --id=mkfastq --csv=Indexes.csv --disable-ui
• cellranger-arc mkfastq --run=BCL --id=mkfastq --csv=Indexes.csv --disable-ui
• cellranger-atac mkfastq --run=BCL --id=mkfastq --csv=Indexes.csv --disable-ui
• bcl2fastq --runfolder-dir BCL --output-dir mkfastq --sample-sheet Indexes.csv

Outputs

• /fastqs: output fastqs
• /logs: mkfastq.out/err and mkfastq.usage

Notes

• memory: "64 GB", cpu: 8, disks: "local-disk {max(BCL*2.5, 96)} SSD"
• throws an error if the disk is >6TB (edit bcl2fastq.wdl to increase cap)
• prints a warning (and proceeds) if the fastq directory already exists
• removes MAKE_FASTQS_CS and Undetermined FASTQs
• the Reports and Stats directories can be found with the fastqs in a folder named after samplesheet

cellranger-count

Inputs

• fastq_path: gs:// path to the fastq folder
• sample: fastq filename prefix to select (specified in the sample sheet supplied to the FASTQ generation software)
• reference: gs:// path to the transcriptome
• technique: "cellranger" or "cellranger-atac"
• lanes (optional): Array[Int] of lanes to subset (default is [], meaning all lanes)
• count_output_path (optional): gs:// path to write outs (default cellranger-count/{basename(fastq_path)})
• log_output_path (optional): gs:// path to write logs (default logs/{basename(fastq_path)})

Commands

• cellranger count --id={id} --transcriptome=reference --fastqs=fastqs --sample={sample} --create-bam=true --include-introns=true --nosecondary --disable-ui
• cellranger-atac count --id={id} --reference=reference --fastqs=fastqs --disable-ui
• rm -rf {id}/SC_RNA_COUNTER_CS

Outputs

• /cellranger-count: output cellranger results
• /logs: count-{id}.out, count-{id}.err, count-{id}.usage

Notes

• memory: "64 GB", cpu: 8, disks: "local-disk {max(fastqs*6+20,128)} SSD"
• throws an error if the disk is >6TB (edit cellranger-count.wdl to increase cap)
• throws an error if the counts directory already exists
• cellranger expects the fastqs to be named as [sample]_S[number]_L00[lane]_[R1/R2/I1/I2]_001.fastq.gz
• the output folder is {id}: equal to {sample} if all lanes are used, {sample}_{lanes} otherwise

spatial-count

Inputs

• fastq_path: gs:// path to the fastq folder
• sample: fastq filename prefix to select (specified in the sample sheet supplied to the FASTQ generation software)
• pucks: array of gs:// paths to each puck
• lanes (optional): Array[Int] of lanes to subset (default is [], meaning all lanes)
• count_output_path (optional): gs:// path to write outs (default spatial-count/{basename(fastq_path)})
• log_output_path (optional): gs:// path to write logs (default logs/{basename(fastq_path)})

Commands

• julia spatial-count.jl fastqs pucks

Outputs

• /spatial-count: SBcounts.h5
• /logs: count-{id}.out, count-{id}.err, count-{id}.usage

Notes

• memory: "{max(fastqs*2.5,64)} GB", cpu: 1, disks: "local-disk {max(fastqs*2.5,64)} SSD"
• throws an error if the disk is >256GB (edit spatial-count.wdl to increase cap)
• WDL expects the fastqs to be named as [sample]_S[number]_L00[lane]_[R1/R2/I1/I2]_001.fastq.gz
• the output folder is {id}: equal to {sample} if all lanes are used, {sample}_{lanes} otherwise

Input→Output

• bcl2fastq: bcls, samplesheets → fastqs
• cellranger-count: fastqs, references → cellranger-count
• spatial-count: fastqs, pucks → spatial-count
• reconstruction: fastqs → reconstruction

Docker Images

us-central1-docker.pkg.dev/velina-208320/docker-bcl2fastq/img:latest
bcl2fastq2 v2.20
Cell Ranger 8.0.1
Cell Ranger ATAC 2.1.0
Cell Ranger ARC 2.0.2

us-central1-docker.pkg.dev/velina-208320/docker-count/img:latest
Python 2.7.16
Python3 3.7.3
Julia 1.8.5
Rust 1.79.0
R 4.3.3

us-central1-docker.pkg.dev/velina-208320/docker-recon/img:latest
Python 3.11
CUDA 12.2

Links

Cell Ranger downloads:
https://www.10xgenomics.com/support/software/cell-ranger/downloads
Cell Ranger ATAC downloads:
https://support.10xgenomics.com/single-cell-atac/software/downloads/latest
Cell Ranger ARC downloads:
https://support.10xgenomics.com/single-cell-multiome-atac-gex/software/downloads/latest

Helpful commands

.h5 commands

• h5dump -n SBcounts.h5: list all contents
• h5dump -d /metadata/num_reads SBcounts.h5: print a specific dataset