rsnt_translations
56,430
edits
GATK: Difference between revisions
Jump to navigation
Jump to search
no edit summary
No edit summary |
No edit summary |
||
| Line 75: | Line 75: | ||
<!--T:13--> | <!--T:13--> | ||
As you probably notice, there are some arguments to be passed directly to | As you probably notice, there are some arguments to be passed directly to Java through the <code>--java-options</code> such as the maximum heap memory (<code>-Xmx8G</code> in the example, reserving 8 Gb of memory for the virtual machine). We recommend that you <b>always</b> use <code>-DGATK_STACKTRACE_ON_USER_EXCEPTION=true</code> since it will give you more information in case the program fails. This information can help you or us (if you need support) to solve the issue. | ||
Note that all options passed to <code>--java-options</code> have to be within quotation marks. | Note that all options passed to <code>--java-options</code> have to be within quotation marks. | ||
=== Considerations regarding our systems === <!--T:50--> | === Considerations regarding our systems === <!--T:50--> | ||
To use GATK | To use GATK on our systems, we recommend that you use the <code>--tmp-dir</code> option and set it to <code>${SLURM_TMPDIR}</code> when in an <code>sbatch</code> job, so that the temporary files are redirected to the local storage. | ||
<!--T:51--> | <!--T:51--> | ||
Also, when using <code>GenomicsDBImport</code> make sure to have the option <code>--genomicsdb-shared-posixfs-optimizations</code> enabled as it "Allow[s] for optimizations to improve the usability and performance for shared Posix Filesystems(e.g. NFS, Lustre)". If not possible or if you are using GNU parallel to run multiple intervals at the same time, please copy your database to <code>${SLURM_TMPDIR}</code> and run it from there as your IO operations might disrupt the function of the filesystem. <code>${SLURM_TMPDIR}</code> is a local storage and therefore is not only faster, but the IO operations would not affect other users. | Also, when using <code>GenomicsDBImport</code>, make sure to have the option <code>--genomicsdb-shared-posixfs-optimizations</code> enabled as it "Allow[s] for optimizations to improve the usability and performance for shared Posix Filesystems(e.g. NFS, Lustre)". If not possible or if you are using GNU parallel to run multiple intervals at the same time, please copy your database to <code>${SLURM_TMPDIR}</code> and run it from there as your IO operations might disrupt the function of the filesystem. <code>${SLURM_TMPDIR}</code> is a local storage and therefore is not only faster, but the IO operations would not affect other users. | ||
===Earlier versions than GATK 4 === <!--T:14--> | ===Earlier versions than GATK 4 === <!--T:14--> | ||