[BioC] summarizeOverlaps mode ignoring inter feature overlaps

[Wei Shi]

On Apr 13, 2013, at 12:08 AM, Martin Morgan wrote:

> On 04/12/2013 06:53 AM, Michael Lawrence wrote:
>> Hi Wei,
>> 
>> summarizeSpliceOverlaps does not yet exist. We held off on that until we
>> determined whether findSpliceOverlaps was useful. And yes,
>> findSpliceOverlaps counts reads on a per-feature basis, so it will double
>> count reads in that way. That's totally intentional (and I'm pretty sure
>> what Thomas was wanting).
> 
> also the object returned can be easily manipulated?
> 
> > olaps = findSpliceOverlaps(galp, genes)
> > olaps
> Hits of length 2
> queryLength: 1
> subjectLength: 2
>  queryHits subjectHits compatible    unique    coding strandSpecific
>   <integer>   <integer>  <logical> <logical> <logical>      <logical>
> 1         1           1      FALSE     FALSE        NA           TRUE
> 2         1           2      FALSE     FALSE        NA           TRUE
> > olaps[mcols(olaps)$unique]
> Hits of length 0
> queryLength: 1
> subjectLength: 2
> 
> and canonically
> 
>  ulaps = olaps[mcols(olaps)$unique]
>  tabulate(subjectHits(ulaps), subjectLength(ulaps))
> 
> see ?Hits
> 
> Martin

But I got no fragments assigned to transcripts. Is this what you meant to do?

>  ulaps = olaps[mcols(olaps)$unique]
>  tabulate(subjectHits(ulaps), subjectLength(ulaps))
[1] 0 0

Cheers
Wei


______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}