[BioC] ChIPpeakAnno settings

[Zhu, Lihua (Julie)]

Sarah,

Please see my comments below.

Best regards,

Julie

On 3/1/11 9:41 AM, "Sheppard, Sarah" <Sarah.Sheppard at umassmed.edu> wrote:

> Hi Julie,
> 
> I would like to use the annotatePeakInBatch function of ChIPpeakAnno to
> identify the distance from a set of peaks to the end of 3'UTRs/transcripts.
> 
> I was going to use these settings, but thought I'd check with you first to see
> if you had any suggestions.
> 
> Peaklist: my own peak data converted from bed format. is it correct that
> strand information can only be as 1/-1?

You can pass +/- for strand

> AnnotationData: I'm using my own supplied data downloaded from BioMart,
> transcript start and end
> output: overlapping
> multiple: true (in case there are multiple 3'UTR models or transcripts, will
> one peak may map to both ensembl models?)

Yes

> maxgap: 100 (does this mean 100bp from the start/end coordinates of the
> transcript?)

Gap 100 means there is a gap of 100 bp (shortest distance) between your peak
and the 3'UTR.

> PeakLocForDistance: I want to use end, because this is where my polyA has been
> cut off- Nathan thinks we should use the middle- can you make a suggestion
> here?

How long is your average peak? You could try both.

> FeatureLocForDistance: end (because I want to know the distance of the peak
> from the end of the transcript)
> 
> 
Sure

> Other options I was thinking about; because the distance to nearest feature
> would then approximately be the length of the 3'UTR
> 
> Annotation Data: coding sequence coordinates
> output- overlapping
> multiple: true
> maxgap- would have to be larger e.g. 1500bp now for the peak to match up to
> the end of the coding sequence, because these technically would not be
> overlapping
> PeakLocForDistance: same issue as above
> FeatureLocForDistance: end
> 
> 
> Annotation Data: 3'UTR coordinates.
> output- both
> multiple: true
> maxgap- 100bp
> PeakLocForDistance: same issue as above
> FeatureLocForDistance: start
> 
> Do you have any suggestions?
> 
> Thanks,
> 
> Sarah