Re: [svtoolkit-help] Regarding Benchmark data and for structural variation other than deletions

[Bob H. <han...@br...>]

On 9/14/12 1:42 AM, Ankita igib wrote:
>
> Hi
>
> My queries are -
>
> (i) I don't understand what are benchmark datasets in svtoolkit folder 
> ? In benchmark folder they have test vcf files. But i have used it for 
> another set of samples, what is benchmark data for them? That's why 
> everytime on running discovery.sh script i got error for differing 
> with benchmark dataset.
The discovery.sh and genotyping.sh scripts are part of the "installation 
test", which does a small test run of the main workflows in Genome 
STRiP, just to make sure the installation is correct.  The benchmark 
data is part of the installation test (i.e. the correct answer).

You can use discovery.sh and genotyping.sh as guides to using Genome 
STRiP, but you should also look at the underlying Q scripts as well to 
understand the steps involved.
And you should be aware of the two most common mistakes:
The -L 1 argument is restricting the analysis to chromosome 1.
In any production use, you should include -reduceInsertSizeDistributions 
in SVPreprocess (you can do this in the installtest too, but it makes 
the test take longer to run; it is a net performance gain on larger data 
sets).
>
> (ii) How can i find/discover duplications even without genotyping them 
> using Genome STRip (as i far i understand Genome STRip also works for 
> genotytping deletions), is there any script available to find 
> duplications using Genome STRip?
Genome STRiP currently does not have a duplication discovery module, 
although we are doing research in this area.
Genotyping duplication sites is also not officially supported.
You might consider CNVnator for duplication discovery based on read depth.

Regards,
-Bob
>
>
>
> -
> Regards
> Ankita
> DST Inspire Fellow (Graduate Student)
> Lab no 604A
> G.N. Ramachandran Knowledge Center for Genome Informatics
> Institute of Genomics and Integrative Biology (CSIR)
> North Campus DU, Near Jubliee Hall
> Delhi - 110007
>
>
> http://igvbrowser.igib.res.in
>
>
>
>
>
>
>
> ------------------------------------------------------------------------------
> Got visibility?
> Most devs has no idea what their production app looks like.
> Find out how fast your code is with AppDynamics Lite.
> http://ad.doubleclick.net/clk;262219671;13503038;y?
> http://info.appdynamics.com/FreeJavaPerformanceDownload.html
>
>
> _______________________________________________
> svtoolkit-help mailing list
> svt...@li...
> https://lists.sourceforge.net/lists/listinfo/svtoolkit-help