[BioC] transcriptLocs2refLocs Bug when strand is negative

[Martin Morgan]

On 08/31/2012 04:22 PM, Cook, Malcolm wrote:
> Herve et al,
>
> I believe that transcriptLocs2refLocs is implemented incorrectly for
> negative strands.
>
> The bug appears in the example of its use you provided in
> https://stat.ethz.ch/pipermail/bioconductor/2011-March/038173.html
>
> I think you will agree that the third result should be:
> [[3]]
> [1] 5899 6291

Hi Malcolm

In the example the exon coordinates (always in 1-based 'leftmost' 
cooredinates') are

start:  c(6291L, 5278L)
   end:  c(6500L, 5899L)

so on the minus strand the 5' -> 3' transcript is from 6500 --> 6291 and 
then 5899 -> 5271, and is 832 nt long.

The first (5') nucleotide on the transcript comes from position 6500, 
the 832nd nucleotide comes from the 3' ('leftmost', when the exon is on 
the minus strand) end of the second exon, at genomic coordinate 5278. 
These are the values that transcriptLocs2refLocs returns.

Martin

>
> As a workaround, I have implemented a version in R using a new generic,
> named `at`
>
> https://gist.github.com/3560692
>
>
> I would be happy to receive confirmation that this is a bug, and comments
> from anyone on the approach to the workaround I have adopted, especially
> as regards performance comparison with
>
> Thanks,
>
> Malcolm Cook
> Stowers Institute for Medical Research
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor
>


-- 
Computational Biology / Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N.
PO Box 19024 Seattle, WA 98109

Location: Arnold Building M1 B861
Phone: (206) 667-2793