J-81/dataset_prep_nf

dataset_prep_nf

Nextflow workflow and associated scripts for preparing protein datasets

Nextflow installation : https://www.nextflow.io/docs/latest/getstarted.html

Uses Nextflow version: 20.07.1

Additional Setup

• IsUnstruct: http://bioinfo.protres.ru/IsUnstruct/
  • Compile and ensure available in either project bin folder or on PATH
• Blast Database (optional):
  • For local blast, you may download and decompress desired database and ensure location specified in configuration file
  • If not performed, this pipeline will download the required database when performing blast locally.
• Configuration File
  • in folder config, path-related parameters will likely need to be changed

Recommend Nextflow Environment Variable Setup

• This will cache all workflow conda environments to a single location. If this is not set, the environment folders will be located in the work directory and not shared between pipelines.

export NXF_CONDA_CACHEDIR="~/SOME/LOCAL/PATH"

• Set version of Nextflow to use

export NXF_VER="20.07.1"

• This will allow monitoring the pipeline using website . While command line output summarizes the pipeline progress as well, tower.nf is much easier to use and can be accessed on any machine with internet access.
  • Follow instructions here: https://tower.nf

Test Run Instructions

• Uses a limited test set of data, should take about 20 minutes for the first time.
• About 15 minutes of first run includes setting up conda environments. If you have set the conda-cache environment, subsequent runs will not require this 15 minute environment setup step.

nextflow run J-81/dataset_prep_nf -profile test

Using Your Own Fasta Files

• The header of the file MUST be the following format:
  • >xxxx_Y
  • where xxxx is the 4 letter PDB ID
  • and Y is the 1 letter Chain ID

Run Instructions

1. Run using defaults of the pipeline, including generating and using a non-redundant scop fasta sequence set.

nextflow run J-81/dataset_prep_nf -r main -with-tower

1. Use defaults of the pipeline with a custom input fasta.

nextflow run J-81/dataset_prep_nf -r main -with-tower --fastaPath {path/to/input/fasta}

1. Run pipeline using custom configuration file.

• Copy the defaults configuration file and modify.
• Location of github: https://github.com/J-81/dataset_prep_nf/raw/main/config/defaults.config

nextflow run J-81/dataset_prep_nf -r main -config {path/to/newconfig/file} -with-tower

Other Useful Parameters

Skip subworkflows, useful for preparing specific parts of a dataset only.

• Skip generating IsUnstruct sequence disorder:

--skipSeqDisorder

• Skip generating secondary structure based switch assignments:

--skipIsSwitch

• Skip generating blast alignment based sequence entropy:

--skipSeqEntropy

• Resume a pipeline where it left off. Useful in cases where an unexpected failure (perhaps internet connection drops causing remote blast to fail) causes the pipeline to fail.

-resume