๐ ๏ธGetting Set Up
Step 0: Requirements
COSAP currently only supports Linux-based operating systems. For other operating systems, consider using COSAP via Docker.
Mapping tasks require at least 16 GB of RAM.
Excluding annotation, data required by COSAP can take up to 50 GB in disk space.
For a typical FASTQ-processing pipeline, it is recommended that you have 3-4 times as much free disk space on your system as the size of your compressed input FASTQ files.
Step 1: Install Miniforge
COSAP requires Miniforge.
Download the script for your system and run it. For example:
bash Miniforge3-Linux-x86_64.shStep 2: Install COSAP
Create a Conda environment with Python 3.9 and COSAP as follows:
mamba create \
--name cosap \
--channel conda-forge \
--channel bioconda \
python=3.9 \
cosapThen activate the environment:
mamba activate cosapStep 3: Download Reference Files
COSAP requires the hg38 reference genome, its indexes, and several known variants sources. The following command will download files necessary for mapping and variant calling:
cosap --download-dataIf you wish, you can also obtain required files yourself and place them at ~/cosap_data or the location you specified for COSAP_LIBRARY_PATH. If you wish for certain files to not be tracked by COSAP, you can specify them in a YAML file placed in the data directory:
ignore:
- file_1.ext
- file_2.extAny files previously present in the data directory will automatically be ignored when the COSAP downloader is run.
Setting up Environment Variables
COSAP_LIBRARY_PATH
Path to directory containing hg38 resources and other required data such as ENSEMBL-VEP cache.
~/cosap_data
COSAP_RAMDISK_PATH
Path to RAM disk directory.
/dev/shm
COSAP_THREADS_PER_JOB
Number of threads usable by tasks. Set to 0 or do not specify for maximum value.
All available
COSAP_IN_MEMORY_MODE
Sets run mode. If true all intermediary files will be kept in memory. Be aware that this mode increase memory consumption significantly.
False
Last updated