Scripts in each module in SMRT pipe?

[metheuse]

I'm new to PacBio data analysis. This question is quite simply/silly. Basically I'm asking how to run secondary analysis by command lines without using SMRT pipe.

The purpose is to detect DNA methylation. PacBio provides a package "SMRTanalysis" to deal with the primary and secondary analysis.
In the secondary analysis, based on its "SMRT pipe reference guide" from this page (http://pacbiodevnet.com/), the protocol is "RS_modificaiton_detection" which includes four modules:
• P_Filter
• P_Mapping
• P_GenomicConsensus
• P_Modification Detection

In each model, I think the program calls multiple individual scripts to accomplish the task. But I can't find any manual online that tells what exactly those scripts are. (If you use SMRTpipe, you don't have to know what scripts each module calls, but I'm trying to run it without Pipe)

The facility here has installed an instance of SMRTanalysis with all its scripts, but only its Portal and View, no Pipe. I'm wondering if anyone knows what scripts are used in each module so that users can run the scripts by themselves instead of using Pipe.

What I found was this page (https://github.com/PacificBioscience...tible-Software) that lists the Compatible Softwares.
I think I can figure out some of them to do the same work as the SMRT "Modification_detection" workflow. Please let me know if the following is correct:
1. Mapping: BLASR
2. GenomicConcensus: Quiver
3. Modification detection: kineticsTools

But I'm missing "filtering". Could anyone tell me which program does "filter"?

Thanks.

[flxlex]

What you could do is set up a run as you desire using the xml file(s) from the portal (for example, setting it up in the portal itself and copying the resulting settings.xml file from the job folder). Collect the path to one or more smrtcell movies in bas.h5 format in a input.fofn file. Then run it with smrtpipe:

fofnToSmrtpipeInput.py input.fofn >input.xml
smrtpipe.py -D TMP=./ -D SHARED_DIR=./ -D NPROC=24 --params=settings.xml xml:input.xml

Once it is finished the output folder will have all the commands it ran, check the workflow folder.

Hope that helps.

[rhall]

SMRT portal runs SMRT pipe in the background, so it would be impossible to have a portal install without having SMRT pipe.
The workflow is not a simple one to one of stages to command line scripts. For example if I go into a workflow directory for a DNA methylation job:

ls ./P_*/*.sh | wc 
68

So without using SMRT pipe you would have to run 68 command line scripts (in the correct order) to get the same results / reports. Granted that is for a distributed task but its still not very practical.
What is stopping you from doing the analysis using SMRT portal? you could then drop down to SMRT pipe for more complex analysis / automated sweeps of parameters etc?

[metheuse]

Thanks for the reply. They are both quite helpful. I'm able to use the pipe now.