Difference between revisions of "PacBio smrtpipe commandline"
From mn/bio/cees-bioinf
| Line 23: | Line 23: | ||
Actual filtering (NPROC is the number of CPUs the job gets): | Actual filtering (NPROC is the number of CPUs the job gets): | ||
<pre>smrtpipe.py -D TMP=./ -D SHARED_DIR=./ -D NPROC=24 --params=desired_settings.xml xml:input.xml &> smrtpipe.err</pre> | <pre>smrtpipe.py -D TMP=./ -D SHARED_DIR=./ -D NPROC=24 --params=desired_settings.xml xml:input.xml &> smrtpipe.err</pre> | ||
| − | === How to obtain the settings.xml file === | + | ==== How to obtain the settings.xml file ==== |
After a smrtportal upgrade, if it is not yet there in the folder /projects/nscdata/scripts/pacbio/smrtpipe_xml_files | After a smrtportal upgrade, if it is not yet there in the folder /projects/nscdata/scripts/pacbio/smrtpipe_xml_files | ||
| Line 36: | Line 36: | ||
Some applications require a reference sequence, e.g. mapping for SNP calling and detecting base modifications, or for running Quiver for assembly polishing. smrtanalysis needs a reference repository, which you can create with this command: | Some applications require a reference sequence, e.g. mapping for SNP calling and detecting base modifications, or for running Quiver for assembly polishing. smrtanalysis needs a reference repository, which you can create with this command: | ||
| − | <pre>referenceUploader -c -p destination_folder -n short_name -f path/to/reference.fasta</pre> | + | <pre>referenceUploader -c -p destination_folder -n short_name -f path/to/reference.fasta</pre> |
| − | NOTES | + | NOTES |
| − | *several references can be added to the same destination_folder | + | *several references can be added to the same destination_folder |
| − | *pick a convenient short name | + | *pick a convenient short name |
| − | *for large fasta files, this process make take some time, as an index file is built | + | *for large fasta files, this process make take some time, as an index file is built |
| − | *when I do this, I get warnings about 'SLF4J', it seems these can be ignored | + | *when I do this, I get warnings about 'SLF4J', it seems these can be ignored |
| − | |||
| − | In the settings.xml file, there is a part saying: | + | '''Using the reference repository''' |
| + | |||
| + | In the ''settings.xml'' file, there is a part saying: | ||
<pre><param name="reference" hidden="true"> | <pre><param name="reference" hidden="true"> | ||
<value>common/references/lambda</value> | <value>common/references/lambda</value> | ||
| − | </param></pre> | + | </param></pre> |
Change the '''common/references/lambda''' part to the full path to the destination_folder<br> | Change the '''common/references/lambda''' part to the full path to the destination_folder<br> | ||
Revision as of 10:32, 8 November 2013
Contents
HOW TO run smrtpipe.py on the command line.
smrtpipe.py is the PacBio pipeline for filtering and mapping PacBio runs, as well as running HGAP. For filtering-only jobs, it splits the raw reads on the adaptor and generates all subreads in fasta or fastq format.
Input:
- input.fofn, list of files to process (see below)
- bax.h5 files (or, before 2.0 software version, bas.h5 files) from the run (one per movie), raw-output from the PacBio
- a settings.xml file, see below, specifies the job. See the end of this page for how to obtain such a xml file
Common steps
Set up environment
module load smrtanalysis/x.y.z # at the time of writing, 'x.y.z' is 2.0.1
Generate smrtcells.fofn file
find path/to/runfile|grep bax.h5 >smrtcells.fofn #NOTE for pre 2,0 versions, use 'bas.h5'!
Convert this file to correct xml format:
fofnToSmrtpipeInput.py smrtcells.fofn >input.xml
copy a settings xml file, e.g. filter_only_settings.xml, HGAP_settings.xml, ...
rsync /projects/nscdata/scripts/pacbio/smrtpipe_xml_files/x.y.z/desired_settings.xml .
NOTE won't work within a screen ..?
Actual filtering (NPROC is the number of CPUs the job gets):
smrtpipe.py -D TMP=./ -D SHARED_DIR=./ -D NPROC=24 --params=desired_settings.xml xml:input.xml &> smrtpipe.err
How to obtain the settings.xml file
After a smrtportal upgrade, if it is not yet there in the folder /projects/nscdata/scripts/pacbio/smrtpipe_xml_files
- In SMRTportal on cod2, set up a job using the protocol you want to run, e.g. RS_Filter_Only, RS_HGAP_Assembly, you don't have to execute it
- note down the job number
- from the folder /projects/nscdata/smrtportal/userdata/jobs/016/, find the folder with your job number
- copy the settings.xml file over to /projects/nscdata/scripts/pacbio/smrtpipe_xml_files/x.y.x/
- give it a smart name (check the other folders), e.g. filter_only_settings.xml, HGAP_settings.xml
Working with a reference for mapping
Some applications require a reference sequence, e.g. mapping for SNP calling and detecting base modifications, or for running Quiver for assembly polishing. smrtanalysis needs a reference repository, which you can create with this command:
referenceUploader -c -p destination_folder -n short_name -f path/to/reference.fasta
NOTES
- several references can be added to the same destination_folder
- pick a convenient short name
- for large fasta files, this process make take some time, as an index file is built
- when I do this, I get warnings about 'SLF4J', it seems these can be ignored
Using the reference repository
In the settings.xml file, there is a part saying:
<param name="reference" hidden="true"> <value>common/references/lambda</value> </param>
Change the common/references/lambda part to the full path to the destination_folder
