BRAKER
BRAKER is a tool for eukaryotic genome annotation. It uses genomic and RNA-Seq data to automatically generate full gene structure annotations in novel genome. BRAKER is based on GeneMark-ET R2 and AUGUSTUS pipelines.
License
Free to use and open source under [Artistic License] (https://opensource.org/licenses/artistic-license-1.0)
Available
Puhti: 2.1.6, 3.0.7
Setting up BRAKER
BRAKER needs some additional setting up steps before using it for the first time.
CSC BRAKER installations do not contain GeneMark or ProtHint softaware packages. While they are free for individual use, their licensing terms do not allow CSC to make a public installation of them. Each user needs to license and install them for their own use.
GeneMark
Go to GeneMark download page, and fill in the form. The version you need is "GeneMark-ES/ET/EP+" for "LINUX 64 kernel 3.10 - 5". Download the program file and the license key. To uncompress the packages:
Copy the uncompressed key file to your home directory.
To tell BRAKER where to find GeneMark, use command line option --GENEMARK_PATH
to point to install location.
BRAKER module contains all the necessary dependencies.
ProtHint
Download and uncompress ProtHint.
wget https://github.com/gatech-genemark/ProtHint/releases/download/v2.6.0/ProtHint-2.6.0.tar.gz
tar xf ProtHint-2.6.0.tar.gz
Use command line option --PROTHINT_PATH
to point to install location.
BRAKER module contains all the necessary dependencies.
AUGUSTUS
AUGUSTUS is included in the installation, but you will need your own copy of AUGUSTUS config directory, as it needs to be writable by the user. You can create this by running command:
It will create directory config
in your current directory.
Use command line option --AUGUSTUS_CONFIG_PATH
to point to the config directory
Usage
In Puhti BRAKER should be used only in batch jobs. Either in normal batch jobs or in interactive batch jobs.
Interactive usage
You can start interactive batch job with command:
BRAKER can utilize several computing cores and can require significant amount of memory so you should reserve more than the default resources for your interactive batch job. For example 4 cores and 32 GB of memory.
In batch job, you can initialize BRAKER environment with command
After that you can launch a BRAKER job with command:
To see the options, run command:
Sample BRAKER command in Puhti:
braker.pl --species=sp1 --genome=Drosophila.dna.fa --prot_seq=Drosophila.pep.fa --prg=gth --trainFromGth --AUGUSTUS_ab_initio --cores=$SLURM_CPUS_PER_TASK --GENEMARK_PATH=/path/to/gmes_linux_64_4 --PROTHINT_PATH=/path/to/ProtHint-2.6.0/bin --AUGUSTUS_CONFIG_PATH /path/to/config
Batch jobs
Sample batch job scrip for BRAKER:
#!/bin/bash
#SBATCH --job-name=BRAKER_Job
#SBATCH --account=project_2012345
#SBATCH --time=24:00:00
#SBATCH --mem=32000
#SBATCH --nodes=1
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=8
#SBATCH --partition=small
# load braker
module load braker
# Use correct paths instead of "/path/to"
braker.pl --species=sp1 --genome=Drosophila.dna.fa --prot_seq=Drosophila.pep.fa \
--prg=gth --trainFromGth --AUGUSTUS_ab_initio --cores=$SLURM_CPUS_PER_TASK \
--GENEMARK_PATH=/path/to/gmes_linux_64_4 \
--PROTHINT_PATH=/path/to/ProtHint-2.6.0/bin \
--AUGUSTUS_CONFIG_PATH /path/to/config
In the batch job example above one task (--ntasks 1) is executed. The BRAKER job uses 8 cores (--cpus-per-task=8 ) with total of 32 GB of memory (--mem=32000). The maximum duration of the job is ten hours (--time 10:00:00 ). All the cores are assigned from one computing node (--nodes=1 ). In the example the project that will be used is project_2012345. This value shuold be replaced by the name of your computing project.
You can submit the batch job file to the batch job system with command:
See the Puhti user guide for more information about running batch jobs.