Hi, Spencer! You have found a little bug in bedToBigBed.
We have fixed it and you can get the new version (2.1) here: http://hgdownload.cse.ucsc.edu/admin/exe/linux.x86_64/bedToBigBed Thanks very much for reporting the problem! -Galt On Mon, Jun 25, 2012 at 12:49 PM, Mahaffey, Spencer < [email protected]> wrote: > Hi Luvina, > > I finally figured out the problem with creating a bigbed file. I have a > bed file with the first 12 fields and I wasn't specifying type or an .as > file. Now I've been able to create a bigBED a couple different ways with > no errors, but with the same problem each time when I go to use them as > tracks. I can get it to create one when I use the .as or just by > specifying the type. > > For example just in case the .as was not quite right I've started just > using the following command line to create the bigBed file. > > ./bedToBigBed -type=bed12 transcripts.allSamples.5s.bed rn4.chrom.sizes > transcripts.allSamples.5.bb > > This seems to work, but the result has one exon. When I convert the > bigBed to bed again, I only have the first blockSize and blockStart value > the rest of the list has been removed. This is true of the values for the > itemRgb column as well. Only the first value in the comma separated list > seems to be in the bigBED file(or its not being output back to the .bed > file). > > Here are the first few lines of the bed file(Created from a gtf file > output by Cufflinks)(I used a slightly modified python script from Wei Li > to create the bed file): > chr1 394971 435563 CUFF.1.1 0 + 394971 435563 255,0,0 2 662,325 0,40267 > chr1 612645 613322 CUFF.2.1 0 . 612645 613322 255,0,0 1 677 0 > chr1 1307207 1307637 CUFF.3.1 2 . 1307207 1307637 255,0,0 1 430 0 > chr1 1421742 1422480 CUFF.4.1 0 . 1421742 1422480 255,0,0 1 738 0 > chr1 1554721 1557379 CUFF.5.1 3 . 1554721 1557379 255,0,0 1 2658 0 > chr1 1615720 1617210 CUFF.6.1 0 . 1615720 1617210 255,0,0 1 1490 0 > chr1 1727887 1728699 CUFF.7.1 0 . 1727887 1728699 255,0,0 1 812 0 > chr1 1922775 1925965 CUFF.8.1 0 . 1922775 1925965 255,0,0 1 3190 0 > chr1 1960722 1961475 CUFF.9.1 0 - 1960722 1961475 255,0,0 2 248,244 0,509 > > I then sorted using bedSort and ran bedToBigBed. > > It just seems like in the bigbed file the comma separated lists have been > truncated to the first value. Here is the bed file output from converting > the bigbed back to bed. > chr1 394971 435563 CUFF.1.1 0 + 394971 435563 255 2 662 0 > chr1 612645 613322 CUFF.2.1 0 . 612645 613322 255 1 677 0 > chr1 1307207 1307637 CUFF.3.1 2 . 1307207 1307637 255 1 430 0 > chr1 1421742 1422480 CUFF.4.1 0 . 1421742 1422480 255 1 738 0 > chr1 1554721 1557379 CUFF.5.1 3 . 1554721 1557379 255 1 2658 0 > chr1 1615720 1617210 CUFF.6.1 0 . 1615720 1617210 255 1 1490 0 > chr1 1727887 1728699 CUFF.7.1 0 . 1727887 1728699 255 1 812 0 > chr1 1922775 1925965 CUFF.8.1 0 . 1922775 1925965 255 1 3190 0 > chr1 1960722 1961475 CUFF.9.1 0 - 1960722 1961475 255 2 248 0 > > I tried to also create the bigBED using type and an .as file. This would > work but only with the following .as file: > table RNAIsoRatBrain > "RNA Isoforms from BNLX/SHRH RNA Seq" > ( > string chrom; "Reference sequence chromosome or scaffold" > uint chromStart; "Transcription start position" > uint chromEnd; "Transcription end position" > string name; "Name of gene" > uint score; "Score" > char[1] strand; "+ or - for strand" > uint thickStart; "Coding region start" > uint thickEnd; "Coding region end" > uint reserved; "Used as itemRgb as of 2004-11-22" > int blockCount; "Number of blocks" > int[blockCount] blockSizes; "Comma separated list of block sizes" > int[blockCount] chromStarts; "Start positions relative to chromStart" > ) > > I could not change reserved to itemRgb and it seems like that should be a > uint[3] but nothing worked. The last column I thought was supposed to be > blockStarts but I just got an error saying it differed from the name > expected chromStarts. So I assume the .as file is not correct but it > wouldn't run other than with this file. I'd either get an error about the > names or the type not matching when I tried to change anything to what I > thought it should be. But the resulting bigBED file also seemed to > truncate the lists for the itemRgb and for the blockSizes and blockStarts > columns. The image looked the same with the file generated with the .as > file. > > I downloaded bedToBigBed again from the website for mac OSX i386. I'm > using OS X 10.6.8. I noticed the linux version has been update more > recently so I thought I would try it on linux. > > Do you have any suggestions about what might cause bedToBigBed to truncate > the lists after the first element? > > Thanks, > Spencer > > > Spencer Mahaffey > Senior Professional Research Assistant > University of Colorado Denver School of Medicine > [email protected]<mailto:[email protected]> > http://phenogen.ucdenver.edu<http://phenogen.ucdenver.edu/> > > > > > _______________________________________________ > Genome maillist - [email protected] > https://lists.soe.ucsc.edu/mailman/listinfo/genome > _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
