The best method for analyzing mobile element insertion sites at some regions by tangram.

Question

[okbys]

Hello,

I'm trying to analyze mobile element insertion sites by gkno fastq-tangram pipeline.

I want to analyze some regions (e.g. all chromosomes) at one time.

However, tangram_detect can analyze only one region at one time.

Any good ideas?

Answer 1

[alistairnward]

I just updated the pipeline to allow specification of a list of regions. Please update to version 1.25.6. If you generate a list of regions (there are a couple of tools within gkno to help with this, e.g. gkno region-by-dict if you want to specify regions of a certain window size), then you can feed them to fastq-tangram using:

gkno pipe fastq-tangram [options] -rls regions -nj 4
gkno pipe fastq-tangram [options] -rlm regions

In both cases, the tool tangram-detect will be called for each of the regions in the list. In the first command, the argument -rls (--region-list-single) supplies the list of regions and all of the tangram-detect calls will appear in a single makefile. Using -nj (--number-jobs) allows multiple jobs to be processed in parallel.

In the second case, the multiple calls to tangram-detect will be placed in multiple makefiles, which can then be submitted manually to a cluster resource for massive parallelisation.

For a tutorial on lists, please see http://blog.gkno.me/post/87905883290/using-lists-for-command-line-arguments.

Please let me know if this solves your problem, or if there is additional support required.

[okbys]

Many thanks!

It's just what I wanted!

Thanks again

Answer 2

[alistairnward]

One more addition to this post. I have updated the pipeline so that you can parallelise alignment of multiple fastq files and then parallelise the detection as part of the same pipeline. A brief tutorial is provided to explain how to go about doing this:

http://blog.gkno.me/post/88719010295/call-meis-on-multiple-fastq-files-v-1-27-1