BioHPC Cloud:
: User Guide

The atac-seq-pipeline has many dependancies. On BioHPC, these dependancies  are implemented in conda as described on its installation instruction. We also keep a copy of the pipeline code in the /programs/atac-seq-pipeline directory.

To run the software, you need to activate the conda virtual environment for  atac-seq-pipeline to access all the dependancies. And copy the pipeline code from  /programs/atac-seq-pipeline to workdir so that you can modify configuration as needed.

Here are the steps to run test data. To run your own data, you need to 

## copy code to workdir, xxxxx is your user ID

cd /workdir/xxxxx            

cp -r /programs/atac-seq-pipeline ./

cd atac-seq-pipeline

## download the test data and test genome

## test data

wget https://storage.googleapis.com/encode-pipeline-test-samples/encode-atac-seq-pipeline/ENCSR356KRQ/ENCSR356KRQ_fastq_subsampled.tar

tar xvf ENCSR356KRQ_fastq_subsampled.tar

## test genome

wget https://storage.googleapis.com/encode-pipeline-genome-data/test_genome_database_hg38_atac.tar

tar xvf test_genome_database_hg38_atac.tar

## as the test run takes a long time to finish, we will run the test within screen

screen

## activate the virtual environment

export PATH=/programs/miniconda3/bin:$PATH

source activate encode-atac-seq-pipeline

## this step is important (replace xxxxx with your user ID):

export PATH=/workdir/xxxxx/atac-seq-pipeline/src:$PATH

## specify the input, and run the job

INPUT=examples/local/ENCSR356KRQ_subsampled.json

java -jar -Dconfig.file=backends/backend.conf cromwell-34.jar run atac.wdl -i ${INPUT} >& run.log &

## once job started, press Ctrl-"a" then press "d" to detach screen, so that job would start in background

## to re-attach to a screen to check status, use the command "screen -r" 

## to process your own data files, you will need to create your json file to specify your data and genome database, use the template: examples/local/ENCSR356KRQ_subsampled.json

(for details about each fields in this json file, see https://github.com/kundajelab/atac-seq-pipeline/blob/master/docs/input.md)

## then run

cd /workdir/xxxxx/atac-seq-pipeline

INPUT=/yourdir/yoursamples.json

java -jar -Dconfig.file=backends/backend.conf cromwell-34.jar run atac.wdl -i ${INPUT} >& run.log &