To me, it seems most sensible then to consider treating the end points as the ranges (width 1). I'm not sure it makes sense to map reads that are apparently aligning to introns. You can map the end points individually and then combine them back together fairly easily. Michael On Wed, Oct 2, 2013 at 11:26 AM, Tim Johnstone wrote: > I am mapping ribosome protected fragments, which may start inside one exon > and end in the following one. I'd like to convert their genomic coordinates > into transcript-based coordinates. > > When I plot the positions of the reads that result from the map function, > I see a lack of reads around each splice junction, so I'm guessing those > reads are being discarded as they're not fully contained in a given exon. > Does findOverlaps consider a read 'within' if it starts in one exon and > finishes in another? It seems like there's something obvious I'm missing. > > I guess [type="start"] is actually what I'm looking for, as otherwise > reads might be double-counted(?), but I still see no option for supplying > that to 'map'. > > Thanks > Tim > > P.S. I was previously using a function I wrote in R, which would run > findOverlaps[type=any], then take all reads that overlapped a given > transcript and convert their positions to transcript coordinates using a > 'refLoc2transcriptLoc' function I wrote. This worked for small sets, > however, it's proving far too memory- and time-inefficient for the larger > transcript sets I am now working with, so map seemed to be a good solution. > > > On Wed, Oct 2, 2013 at 1:55 PM, Michael Lawrence < > lawrence.michael@gene.com> wrote: > >> What do you want to happen in that case? Those reads would be outside the >> annotated transcript. >> >> >> On Wed, Oct 2, 2013 at 10:48 AM, Tim Johnstone < >> timothy.johnstone@yale.edu> wrote: >> >>> Thanks for your response! >>> >>> From what I can see, map calls 'findOverlaps' using [type="within"]. Is >>> it possible to supply an argument to 'map' such that it will call >>> findOverlaps with [type="any"]? I am running map(GappedAlignments, >>> GRangesList), where the GRangesList is a list of exons, and I need to >>> include reads that overlap the exons at any point, but it currently looks >>> like I'm not capturing the reads at intron-exon boundaries as they don't >>> fully overlap. >>> >>> Thanks, >>> Tim J. >>> >>> >>> On Tue, Sep 17, 2013 at 1:06 PM, Michael Lawrence < >>> lawrence.michael@gene.com> wrote: >>> >>>> There is a method on the map generic in GenomicRanges that will do this >>>> given a GRanges of genomic/global coordinates, and a GRangesList defining >>>> transcripts or any other type of compound region. The returned >>>> RangesMapping object tells you the hits (which overlapped which) and for >>>> each hit the local coordinates. >>>> >>>> Michael >>>> >>>> >>>> On Tue, Sep 17, 2013 at 9:33 AM, Tim Johnstone < >>>> timothy.johnstone@yale.edu> wrote: >>>> >>>>> To add to my previous email, I just found the refLocsToLocalLocs >>>>> function in the VariantAnnotation package you co-authored, but need to >>>>> convert single positions that align anywhere in the transcript, rather than >>>>> ranges that align to just the coding region. It looks like it may be >>>>> possible to get this behavior by supplying exonsByTx as both the second and >>>>> the third arguments, but let me know if there is an easier way. >>>>> >>>>> Best, >>>>> Tim Johnstone >>>>> >>>>> >>>>> On Tue, Sep 17, 2013 at 12:19 PM, Tim Johnstone < >>>>> timothy.johnstone@yale.edu> wrote: >>>>> >>>>>> Dear Michael, >>>>>> >>>>>> I am working with the GenomicRanges bioconductor package, and I've >>>>>> been trying to find a way to convert a set of locations in genomic >>>>>> coordinates to transcript coordinates. I found an old mailing list >>>>>> conversationin which you indicated you were planning to implement that function >>>>>> (refLocs2TranscriptLocs or globalToLocal) in GenomicRanges/GenomicFeatures, >>>>>> but in looking through the source code I can't find it. Was this function >>>>>> ever implemented, and if so, would it be possible for you to send it to me? >>>>>> >>>>>> Thanks! >>>>>> Tim Johnstone >>>>>> >>>>>> -- >>>>>> Timothy G. Johnstone >>>>>> Yale University School of Medicine >>>>>> New Haven, CT 06510 >>>>>> 503-318-0962 >>>>>> >>>>> >>>>> >>>>> >>>>> -- >>>>> Timothy G. Johnstone >>>>> Yale University School of Medicine >>>>> New Haven, CT 06510 >>>>> 503-318-0962 >>>>> >>>> >>>> >>> >>> >>> -- >>> Timothy G. Johnstone >>> Yale University School of Medicine >>> New Haven, CT 06510 >>> 503-318-0962 >>> >> >> > > > -- > Timothy G. Johnstone > Yale University School of Medicine > New Haven, CT 06510 > 503-318-0962 > [[alternative HTML version deleted]]