Skip to content

Introduction to HCBC pipelines

Content -

Data Management

AWS

  • Samplesheet input files for pipelines
    • pipelineName_PI_hbcNNNNNN
    • Have a copy in project folder in O2
    • Manually removing weekly during platform meeting
  • Raw data is under input folder
    • See instructions below to move data in/out
    • pipelineName_PI_hbcNNNNNN
    • lifecycle 14 days
  • Pipeline outputs are under results:
    • pipelineName_PI_hbcNNNNNN
    • lifecycle 14 days for bigger than 1gb
    • Move output pipeline to project folder under final folder

Move data in/out of AWS

Follow this to copy data in and out of our AWS space:

  • Log in into transfer node in O2
  • Type sudo -su bcbio to be bcbio user
  • Use this command to copy data to AWS: 
    /usr/local/bin/aws s3 sync $FOLDER_WITH_FASTQ s3://hcbc-seqera/input/rnaseq_piname_hbcNNNN
  • Use this command to copy data from AWS: 
    /usr/local/bin/aws s3 sync s3://hcbc-seqera/results/rnaseq_piname_hbcNNNN $FOLDER_PROJECT
    Make sure bcbio group has read/write access to the folders otherwise aws command won't work, but won't error either.

Parameters

RNAseq

  • We use salmon with bam files produced by STAR mapped to transcriptome for quantification

CHIPseq

  • It can analyze multiple antibodies in one pipeline run (pipeline splits samples by antibody)
  • Default parameters
  • de-duplication for all samples
  • bowtie is set up with these extra parameters: --sensitive-local -X 1000 (this is only true in seqera dev environment, not production)
  • macs_gsize needs to be setup for each species accordingly tools

CUT&RUN

  • Run once per antibody (because pipeline does not split samples by antibody)
  • In the samplesheet, if the number of samples in an antibody group is the same as the number of samples in its control group, the pipeline will match individual antibody samples to individual control samples using the replicate number. If the number of samples in the antibody group is not the same as the number of samples in the control group, then replicate 1 of the control group will be used as the control sample for all antibody samples. You may not desire this behavior. If not, you may consider selecting an optimal control for all samples in an antibody group and either 1) using this optimal control for all samples in the antibody group or 2) using paired controls when available and using this optimal control sample otherwise.
  • Turn on dedup_target_reads
  • Use both macs2 and seacr for peakcalling (list macs2 first so it is used as primary)
  • Normalization mode is set to CPM (can be changed if client has spike-in samples)
  • Depending on the number of samples, user may want to skip deeptools processes involving all samples
  • processes including SAMTOOLS_SORT, BEDTOOLS_SORT, SAMTOOLS_CUSTOMVIEW, FRAG_LEN_HIST, and DEEPTOOLS_PLOTHEATMAP_GENE_ALL are given more memory than nf-core default

ATACseq

All peaks nf-core-atac-seq_shift: - shift is on - keep_dup is false

NFR peaks nf-core-atac-seq_shift_NFR: - same than previous except parameters for Aligmentsieve: - --minFragmentLength 0 - --maxFragmentLength 120

Note: Recommendation to check the fragment length distribution after the run to make sure you're capturing the NFRs. Note: We don't need the MN(180, 247), DN (315, 473) and TN (558, 615), unless it's a specific case where we are looking at global shifts in accessibility

Nextflow in Seqera platform

  • Create an user here: https://cloud.seqera.io/login
  • Ask Platform team to add you to HCBC workspace
  • Transfer data to HCBC S3: Ask Alex/Lorena. Files will be at our S3 bucket input/pipelineName_PI_hbcNNNNNN folder

RNAseq

  • Prepare the CSV file according this instructions. File should look like this:
sample,fastq_1,fastq_2,strandedness
CONTROL_REP1,s3path/AEG588A1_S1_L002_R1_001.fastq.gz,s3path/AEG588A1_S1_L002_R2_001.fastq.gz,auto
CONTROL_REP1,s3path/AEG588A1_S1_L003_R1_001.fastq.gz,s3path/AEG588A1_S1_L003_R2_001.fastq.gz,auto
CONTROL_REP1,s3path/AEG588A1_S1_L004_R1_001.fastq.gz,s3path/AEG588A1_S1_L004_R2_001.fastq.gz,auto

Use bcbio_nfcore_check(csv_file) to check the file is correct.

You can add more columns to this file with more metadata, and use this file as the coldata file in the templates.

  • Safe the file under meta folder
  • Upload this file to our Datasets in Seqera using the name of the project but starting with pipelineName_PI_hbcNNNNNN
  • Go to Launchpad, select nf-core_rnaseq pipeline, and select the previous created Datasets in the input parameter after clicking in Browser
  • Select an output directory with the same name used for the Dataset inside the results/pipelineName_PI_hbcNNNNNN folder in S3
  • When pipeline is done, data will be copied to our on-premise HPC in the scratch system under scratch/groups/hsph/hbc/bcbio/ folder

Nextflow in O2

  • Nextflow is available at /n/app/bcbio/nextflow/nextflow.
  • Singularity containers at available at /n/app/singularity/containers/shared/bcbio/.
  • Cluster config: /n/app/bcbio/nextflow/o2.config

An example of sbatch script is:

#!/bin/bash

#SBATCH --job-name=Nextflow      # Job name
#SBATCH --partition=priority            # Partition name
#SBATCH --time=1-23:59                 # Runtime in D-HH:MM format
#SBATCH --nodes=1                      # Number of nodes (keep at 1)
#SBATCH --ntasks=1                     # Number of tasks per node (keep at 1)
#SBATCH --cpus-per-task=1            # CPU cores requested per task (change for threaded jobs)
#SBATCH --mem=12G                     # Memory needed per node (total)
#SBATCH --error=jobid_%j.err           # File to which STDERR will be written, including job ID
#SBATCH --output=jobid_%j.out          # File to which STDOUT will be written, including job ID
#SBATCH --mail-type=ALL                # Type of email notification (BEGIN, END, FAIL, ALL)

module load java/jdk-21.0.2
export NXF_APPTAINER_CACHEDIR=/n/app/singularity/containers/shared/bcbio/nf-core-rnaseq-3.14.0
export NXF_SINGULARITY_LIBRARYDIR=/n/app/singularity/containers/shared/bcbio/nf-core-rnaseq-3.14.0

/n/app/bcbio/nextflow/nextflow run nf-core/rnaseq -r 3.14.0 -profile singularity \
    -c /n/app/bcbio/nextflow/o2.config -c /n/app/bcbio/nextflow/rnaseq.resources.config \
    -params-file /n/app/bcbio/nextflow/rnaseq.json \
    --input samplesheet.csv --outdir this_folder -resume

RNAseq

Containers at /n/app/singularity/containers/shared/bcbio/nf-core-rnaseq-3.14.0

viralrecon

Read documentation here.

This is an example for test data:

module load java/jdk-21.0.2
export NXF_APPTAINER_CACHEDIR=/n/app/singularity/containers/shared/bcbio/nf-core-viralrecon_2.6.0
export NXF_SINGULARITY_LIBRARYDIR=/n/app/singularity/containers/shared/bcbio/nf-core-viralrecon_2.6.0

/n/app/bcbio/nextflow/nextflow run nf-core/viralrecon -r 2.6.0 -profile singularity,test --outdir this_folder -resume
To run your data, prepare input file following this doc, and run it like this:

/n/app/bcbio/nextflow/nextflow run nf-core/viralrecon -r 2.6.0 -profile singularity --outdir this_folder --input samplesheet.csv -resume

Nextflow in FAS

We use FAS to run pipelines in scratch, main storage location for data is still O2.

  • Run pipelines in scratch: /n/netscratch/hsph_bioinfo/Lab
  • Keep downstream analysis in PIs folder: /n/holylfs05/LABS/hsph_bioinfo/Lab/PIs

Pipelines that have been run so far in FAS: - rnaseq - scrnaseq - cutandrun

Before using nextflow, you need to load a recent version of java:

module load jdk/21.0.2-fasrc01

Use nextflow at /n/holylfs05/LABS/hsph_bioinfo/Lab/shared_resources/nextflow/nextflow

Use config file at /n/holylfs05/LABS/hsph_bioinfo/Lab/shared_resources/nextflow/cannon.config

Example command to run in an interactive job: Note that this is only for test datasets requiring minimal CPUs, memory, and parallelization. Otherwise, submit as sbatch. 

/n/holylfs05/LABS/hsph_bioinfo/Lab/shared_resources/nextflow/nextflow run nf-core/rnaseq -profile test,singularity --outdir tmp -c /n/holylfs05/LABS/hsph_bioinfo/Lab/shared_resources/nextflow/fas.config

For non-test data, you will use sbatch to submit the head job of the pipeline, which will in turn submit the child jobs. Modify this template as needed before using it:

#!/bin/bash

#SBATCH --job-name=Nextflow      # Job name
#SBATCH --partition=shared            # Partition name
#SBATCH --time=0-48:59                 # Runtime in D-HH:MM format
#SBATCH --nodes=1                      # Number of nodes (keep at 1)
#SBATCH --ntasks=1                     # Number of tasks per node (keep at 1)
#SBATCH --mem=16G                     # Memory needed per node (total)
#SBATCH --error=jobid_%j.err           # File to which STDERR will be written, including job ID
#SBATCH --output=jobid_%j.out          # File to which STDOUT will be written, including job ID
#SBATCH --mail-type=ALL                # Type of email notification (BEGIN, END, FAIL, ALL)

module load jdk/21.0.2-fasrc01

# modify these paths as necessary to point to the containers for the pipeline you're using
export NXF_APPTAINER_CACHEDIR=/n/holylfs05/LABS/hsph_bioinfo/Lab/shared_resources/nextflow/nfcore-rnaseq
export NXF_SINGULARITY_LIBRARYDIR=/n/holylfs05/LABS/hsph_bioinfo/Lab/shared_resources/nextflow/nfcore-rnaseq

# Optional: if you'd like to monitor your run in Seqera Platform, set up a token there and use it here
export TOWER_WORKSPACE_ID=268530979103043
export TOWER_ACCESS_TOKEN=<your_access_token>

OUTPUT=path_to_results

/n/holylfs05/LABS/hsph_bioinfo/Lab/shared_resources/nextflow/nextflow run nf-core/rnaseq
  -r 3.14.0 \
  -profile singularity \
  # next line is for passing pipeline parameters. can also insead use --params-file and use JSON downloaded from seqera
  -c analysis.config \
  # next line is for optimizing resource requests, file does not yet exist for all pipelines
  -c rnaseq.resources.config \ 
  --outdir $OUTPUT
  -c cannon.config \
  # next line optional, if you want to monitor your run in Seqera Platform
  -with-tower
  -resume