Skip to content

Pipeline Tutorial

Jump to bottom
Samantha edited this page Nov 22, 2021 · 17 revisions

Welcome to the iCLIP Pipeline Tutorial!

Getting Started

Review the information on the WikiPage for a complete overview the pipeline. Use the information below to run specific tests.

  1. Change working directory to the iCLIP repository
cd /data/RBL_NCI/Pipelines/iCLIP/[version number]
  1. Initialize Pipeline
sh run_snakemake.sh -p initialize -o /path/to/output/dir

Select a test set

Determine the type of test you would like to complete, and choose from the list below.

  1. Small test (multiplex_flag="N", splice_aware="N", DE_method="none")

    test_1: one sample, hg38

  2. Full test (multiplex_flag="Y", splice_aware="Y", DE_method="none")

    test_2: one sample, mm10

  3. Full test (multiplex_flag="Y", splice_aware="Y", DE_method="MANORM")

    test_3: one sample, hg38

  4. Full test (multiplex_flag="Y", splice_aware="Y", DE_method="MANORM")

    test_4: one sample, hg38

Copy the necessary files to your output directory.

Usage: /data/CCBR_Pipeliner/iCLIP/test/run_test.sh -t test_id
	-t options: test_1, test_2, test_3, test_4
Usage:  -o output_dir
	-o path to output directory

For example, to run test_1:

sh /data/CCBR_Pipeliner/iCLIP/test/run_test.sh -t test_1 -o /path/to/output/dir

Preparing Configs

Config files have been pre-populated for the majority of variables.

  1. Open the /path/to/output/dir/config/snakemake_config.yaml
  2. Edit the outputDir to your /path/to/output/dir
#path to output directory
outputDir: "/path/to/output/dir/"
  1. Save your changes

Run the pipeline

  1. Move to the iCLIP dir
cd /data/RBL_NCI/Pipelines/iCLIP/[version number]
  1. Check the manifests to ensure here are no errors.
#run command
sh run_snakemake.sh -p check -o /path/to/output/dir/

#ensure the clean_manifest file was created
cat /path/to/output/dir/qc/manifest_check.txt

NOTE: If there are no errors, and you have not selected a DE_method, the manifest file will be empty. If there are errors, you will see them listed. If you chose a DE_method, you will see a comparative summary to review.

  1. Complete a dry-run to troubleshoot any initial issues
sh run_snakemake.sh -p dry -o /path/to/output/dir/
  1. Execute pipeline on the cluster OR locally
#submit to the cluster (recommended)
sh run_snakemake.sh -p cluster -o /path/to/output/dir/

#run on local interactive node
sh run_snakemake.sh -p local -o /path/to/output/dir/

Review outputs

  • Review the expected outputs on the WikiPage.

Troubleshooting

  • Review the steps for troubleshooting on the WikiPage.
Clone this wiki locally