On Wed, Nov 25, 2009 at 5:59 AM, Johannes Waage <johannes.waage@gmail.com>wrote: > Thank you for your reply. > > Trying to pass blockSizes as a column with i.e. "20,20", disrupts the > interaction with UCSC. My RangedData looks like this. > > RangedData with 24 rows and 9 value columns across 1 space > space ranges | itemRgb score name > thickStart thickEnd blockCount blockSizes blockStarts strand > <character> <iranges> | <character> <numeric> <character> > <numeric> <numeric> <numeric> <character> <character> <factor> > 1 chr2 [5971505, 5972574] | 0,0,0 1000 > Upf2 0 0 2 20,20 0,1049 + > 2 chr2 [5872579, 5878536] | 0,0,0 1000 > Upf2 0 0 2 20,20 0,5937 + > 3 chr2 [5872579, 5955177] | 150,0,0 1000 > Upf2 0 0 2 20,20 0,82578 + > 4 chr2 [5950571, 5955177] | 0,0,0 1000 > Upf2 0 0 2 20,20 0,4586 + > 5 chr2 [5872579, 5894534] | 150,0,0 1000 > Upf2 0 0 2 20,20 0,21935 + > 6 chr2 [5882738, 5894534] | 0,0,0 1000 > Upf2 0 0 2 20,20 0,11776 + > 7 chr2 [5948514, 5950427] | 0,0,0 1000 > Upf2 0 0 2 20,20 0,1893 + > 8 chr2 [5947344, 5948278] | 0,0,0 1000 > Upf2 0 0 2 20,20 0,914 + > 9 chr2 [5961587, 5967963] | 0,0,0 1000 > Upf2 0 0 2 20,20 0,6356 + > 10 chr2 [5909498, 5924610] | 0,0,0 1000 > Upf2 0 0 2 20,20 0,15092 + > > Removing bockSizes and blockStarts from the object allows upload, but of > course doesn't get me what I need. It seems the confusion here stems from the differences between the IRanges and BED interval bound conventions. Ranges objects are always fully closed intervals, 1-based, while BED is 0-based and half-open. Thus, an adjustment is made for the chromStarts/chromEnds (which become a Ranges object), while no such translation occurs for the textual blockStarts. There is an adjustment when retrieving the blocks as a Ranges object via the blocks() function. In your case, you have correctly specified the blockStarts as 0-based, however the Ranges shown are 1-based and thus will be decremented when passed to UCSC, causing the blocks to extend outside the feature range. I guess rtracklayer needs to adjust those textual blockStarts. Starting to wonder if it wouldn't make more sense to store those numbers as an IntegerList, rather than CSV. > Furthermore, how do I enable colors as per the preamble itemRgb=On in > custom tracks? > > There are several ways. One way: when you upload the track you can specify itemRgb=TRUE. You can also modify the 'trackLine' slot after coercing the RangedData to a UCSCData. > Thank you in advance, > > Johannes > > > On Mon, Nov 23, 2009 at 3:31 PM, Michael Lawrence <mflawren@fhcrc.org>wrote: > >> >> >> On Mon, Nov 23, 2009 at 2:08 AM, Johannes Waage <johannes.waage@gmail.com>> > wrote: >> >>> Hi all, >>> >>> Rtracklayer interacts nicely with UCSC GB, but does anyone know if there >>> is >>> support for uploading and displaying BED-tracks with thin and thick lines >>> (e.g. for visualizing splice-junction data from RNA-seq). >>> Is it possible to manipulate thickStart and thickEnd as well as other >>> optional BED-fields? >>> >>> >> All BED fields are supported. If your RangedData has the thickStart and >> thickEnd columns, they will be transmitted to the browser, as long as you're >> using the BED format for the transmission. By default, the format is >> automatically determined. Basically, BED is only used when there is no >> numeric score column, otherwise WIG or bedGraph is chosen. To force BED, >> pass format = "bed" when uploading the track. >> >> >>> Thanks in advance, >>> >>> JW, >>> Biotech Researh and Innovation Center >>> Uni. of Copenhagen >>> >>> [[alternative HTML version deleted]] >>> >>> _______________________________________________ >>> Bioconductor mailing list >>> Bioconductor@stat.math.ethz.ch >>> https://stat.ethz.ch/mailman/listinfo/bioconductor >>> Search the archives: >>> http://news.gmane.org/gmane.science.biology.informatics.conductor >>> >> >> > [[alternative HTML version deleted]]