[BioC] QuasR for miRNA

Michael Stadler michael.stadler at fmi.ch
Tue May 20 08:40:00 CEST 2014


Dear Simon,

I hope you don't mind that I am copying the Bioconductor list - your
question may be relevant to others as well.

The reason for extending the mature miRNA coordinates depending on their
strand lies in the way QuasR's qCount determines if an alignment
overlaps with a query region:

By default, qCount counts alignments if the 5'-end base is contained in
the query region (see ?qCount, selectReadPosition="start"). The example
in the vignette (section 6.3.3) therefore constructs a query region for
each mature miRNA around its 5'-end, which corresponds to
start(matureMirs) for plus-strand miRNAs, and to end(matureMirs) for
minus-strand miRNAs.

The resulting query region is 7 nucleotides long and allows for some
variability in the processing of mature miRNAs, yet it is narrow enough
to not also count reads that are unlikely to be Dicer-processed mature
miRNAs.

You are right that in the vignette, only mature miRNAs are quantified.
This is just an illustration; you could also create query regions
corresponding to the pre-miRNAs to also include reads that stem from the
precursor.

Michael

On 19.05.2014 18:12, Simon de Bernard wrote:
> Dear Michael,
> 
> I was looking at the miRNA part of your vignette for QuasR (http://master.bioconductor.org/packages/release/bioc/vignettes/QuasR/inst/doc/QuasR.pdf).
> 
> When you extend the the start and end positions (page 34), should you really test for the strand?
> Shouldn't it simply be:
>  
> start(matureMirsExtended) <- start(matureMirs) - 3
> end(matureMirsExtended) <- end(matureMirs) + 3
> 
> Also, in your exemple, you do not extend the preMirs. Is this intentional?
> 
> Thanking you in advance,
> 
> Best regards,
> 
> Simon.
>



More information about the Bioconductor mailing list