Skip to content

Megahit

Megahit is an ultra-fast assembly tool for metagenomics data.

License

Free to use and open source under GNU GPLv3.

Available

  • Puhti: 1.2.9

Usage

On Puhti, Megahit is activated by loading the biokit environment:

module load biokit

For usage help, use command:

megahit -h

Assembling metagenomic data can be very resource demanding. Note that you should not run Megahit on the login nodes of Puhti. For any real analysis task, we recommend running Megahit as a batch job.

Sample Megahit batch job:

#!/bin/bash
#SBATCH --job-name=Megahit
#SBATCH --account=<project>
#SBATCH --time=12:00:00
#SBATCH --ntasks=1
#SBATCH --nodes=1
#SBATCH --output=megahit_out_8
#SBATCH --error=megahit_err_8
#SBATCH --cpus-per-task=8
#SBATCH --mem=32G
#SBATCH --partition=small

module load biokit
srun megahit -1 reads_1.fastq -2 reads_2.fastq -t $SLURM_CPUS_PER_TASK --m 32000000000 -o result_directory

In the example above <project> should be replaced with your project name. You can use csc-projects to check your CSC projects. Maximum running time is set to 12 hours (--time=12:00:00). As Megahit uses thread-based parallelization, the process is considered as one job that should be executed within one node (--ntasks=1, --nodes=1). The job reserves eight cores (--cpus-per-task=8) that can use in total up to 32 GB of memory (--mem=32G). Note that the number of cores to be used needs to be defined in actual Megahit command too. That is done with Megahit option -t. In this case we use $SLURM_CPUS_PER_TASK variable that contains the --cpus-per-task value (we could as well use -t 8, but then we have to remember to change the value if the number of the reserved CPUs is changed).

The job is submitted to the batch job system with sbatch command. For example, if the batch job file is named as megahit_job.sh, then the submission command is:

sbatch megahit_job.sh

More information about running batch jobs can be found from the batch job section of the Puhti user guide.

More information