Bureaucrats, cc_docs_admin, cc_staff
2,879
edits
Parabricks: Difference between revisions
Jump to navigation
Jump to search
Marked this version for translation
(remove draft tag, mark for translation) |
(Marked this version for translation) |
||
| Line 3: | Line 3: | ||
<translate> | <translate> | ||
<!--T:1--> | |||
Parabricks is a software suite for performing secondary analysis of next generation sequencing (NGS) DNA data. Parabricks is extremely fast: It can analyze the whole human genome in about 45 minutes, compared to about 30 hours for 30x [https://en.wikipedia.org/wiki/Whole-genome_shotgun WGS] data. It achieves this performance through tight integration with GPUs. | Parabricks is a software suite for performing secondary analysis of next generation sequencing (NGS) DNA data. Parabricks is extremely fast: It can analyze the whole human genome in about 45 minutes, compared to about 30 hours for 30x [https://en.wikipedia.org/wiki/Whole-genome_shotgun WGS] data. It achieves this performance through tight integration with GPUs. | ||
<!--T:2--> | |||
You can learn more at [http://www.nvidia.com/parabricks www.nvidia.com/parabricks] | You can learn more at [http://www.nvidia.com/parabricks www.nvidia.com/parabricks] | ||
=Usage in Compute Canada Clusters = | =Usage in Compute Canada Clusters = <!--T:3--> | ||
<!--T:4--> | |||
'''This software was provided freely by NVidia to help with research on COVID19 until Sunday, 17 May 2020.''' | '''This software was provided freely by NVidia to help with research on COVID19 until Sunday, 17 May 2020.''' | ||
Since this free period has expired, you must have your own license arrangement with NVidia in order | Since this free period has expired, you must have your own license arrangement with NVidia in order | ||
to use Parabricks on Compute Canada equipment. | to use Parabricks on Compute Canada equipment. | ||
== Finding and loading Parabricks == | == Finding and loading Parabricks == <!--T:5--> | ||
<!--T:6--> | |||
Parabricks can be looked for as a regular module through module spider: | Parabricks can be looked for as a regular module through module spider: | ||
{{Command | {{Command | ||
| Line 20: | Line 24: | ||
}} | }} | ||
<!--T:7--> | |||
Likewise, it can be loaded through LMOD modules: | Likewise, it can be loaded through LMOD modules: | ||
<!--T:8--> | |||
{{Command | {{Command | ||
|module load parabricks/2.5.0 | |module load parabricks/2.5.0 | ||
| Line 27: | Line 33: | ||
== Example of use == | == Example of use == <!--T:9--> | ||
<!--T:10--> | |||
Before you use Parabricks, make sure you have gone through the [https://www.nvidia.com/en-us/docs/parabricks/ Parabricks documentation], including their standalone tools and pipelines. Also make sure you know [https://docs.computecanada.ca/wiki/Using_GPUs_with_Slurm how to request GPUs in Compute Canada clusters]. Once you understand the above, you can submit a job like: | Before you use Parabricks, make sure you have gone through the [https://www.nvidia.com/en-us/docs/parabricks/ Parabricks documentation], including their standalone tools and pipelines. Also make sure you know [https://docs.computecanada.ca/wiki/Using_GPUs_with_Slurm how to request GPUs in Compute Canada clusters]. Once you understand the above, you can submit a job like: | ||
<!--T:11--> | |||
<pre> | <pre> | ||
#!/bin/bash | #!/bin/bash | ||
| Line 40: | Line 48: | ||
#SBATCH --time=5:00:00 | #SBATCH --time=5:00:00 | ||
<!--T:12--> | |||
module load parabricks/2.5.0 | module load parabricks/2.5.0 | ||
<!--T:13--> | |||
DATA_DIR=/path/to/data | DATA_DIR=/path/to/data | ||
OUT_DIR=/path/to/output | OUT_DIR=/path/to/output | ||
| Line 54: | Line 64: | ||
</pre> | </pre> | ||
<!--T:14--> | |||
{{Note | {{Note | ||
|Make the path to the files absolute real paths (i.e. with the command <code>realpath .</code>) | |Make the path to the files absolute real paths (i.e. with the command <code>realpath .</code>) | ||
}} | }} | ||
== Common issues == | == Common issues == <!--T:15--> | ||
=== Almost immediate failure === | === Almost immediate failure === <!--T:16--> | ||
<!--T:17--> | |||
If your first test fails right away, there might be a missing module or some environmental variable clash. To solve this try: | If your first test fails right away, there might be a missing module or some environmental variable clash. To solve this try: | ||
<!--T:18--> | |||
{{Command | {{Command | ||
|module --force purge | |module --force purge | ||
| Line 71: | Line 84: | ||
}} | }} | ||
=== Later failure === | === Later failure === <!--T:19--> | ||
<!--T:20--> | |||
Often Parabricks may not give you a clear traceback of the failure. This usually means that that you did not request enough memory. If you are reserving a full node already through <code>--nodes=1</code>, we suggest you also use all the memory in the node with <code>--mem=0</code>. Otherwise, make sure that your pipeline has enough memory to process your data. | Often Parabricks may not give you a clear traceback of the failure. This usually means that that you did not request enough memory. If you are reserving a full node already through <code>--nodes=1</code>, we suggest you also use all the memory in the node with <code>--mem=0</code>. Otherwise, make sure that your pipeline has enough memory to process your data. | ||
== Hybrid usage == | == Hybrid usage == <!--T:21--> | ||
<!--T:22--> | |||
Parabricks uses both CPU and GPUs. During our tests, Parabricks used at least 10 CPUs, so we recommend to ask for at least that amount through <code>--cpus-per-task=10</code> | Parabricks uses both CPU and GPUs. During our tests, Parabricks used at least 10 CPUs, so we recommend to ask for at least that amount through <code>--cpus-per-task=10</code> | ||
=References = | =References = <!--T:23--> | ||
[http://www.nvidia.com/parabricks Parabricks Home] | [http://www.nvidia.com/parabricks Parabricks Home] | ||
<!--T:24--> | |||
[[Category:Bioinformatics]] | [[Category:Bioinformatics]] | ||
[[Category:Software]] | [[Category:Software]] | ||