Hi again Janina, The fix I described below is on our main site now, but we found another, more subtle bug for some cases of Table Browser intersections with bigWig tracks.
If you are using a bigWig track as your primary table (the first table you select in the Table Browser) and creating an intersection with another table, and you are using the "Base-pair-wise intersection (AND)" option to do the intersection, the results might be missing some bases. (Specifically, the results will be missing items from the bigWig track that were not completely overlapped by items from the second table.) This second problem was fixed on our test site in the last few hours, so we recommend continuing to use the test site until the bug fix has made it to the main site (in about 3 weeks). Once again the test site is http://genome-test.cse.ucsc.edu. You may want to double-check any results you have gotten from the test site for the last few days. Please let us know if you have any questions or concerns. -- Brooke Rhead UCSC Genome Bioinformatics Group On 07/01/11 14:45, Brooke Rhead wrote: > Hi Janina, > > Thank you for sending an example. You helped us track down a bug in the > Genome Browser: intersections in the Table Browser were not working > correctly when the table selected was a bigWig type file (described in > this paper, if you are interested: > http://bioinformatics.oxfordjournals.org/content/26/17/2204.long). > > We have fixed the problem, and the fix should be out on our regular site > next week. In the meantime, you can use our test site to intersect > custom tracks with bigWig tracks: http://genome-test.cse.ucsc.edu. > > Be aware that there is a lot of data on the test site that has not > undergone our quality assurance process and is subject to change. > > Thanks again for bringing this bug to our attention. If you have > further questions or problems, please feel free to write back to > [email protected]. > > -- > Brooke Rhead > UCSC Genome Bioinformatics Group > > > On 06/30/11 11:49, Brooke Rhead wrote: >> Hi Janina, >> >> I think the difference you are seeing is probably due to a different >> file format being present in the 2009 (hg19) version. Can you give us >> an example (the group, track, and table from the Table Browser) that fails? >> >> -- >> Brooke Rhead >> UCSC Genome Bioinformatics Group >> >> >> On 06/29/11 06:49, Janina Goernemann wrote: >>> Hello, >>> I have a problem with the "intersection" function in the table >> browser, but only with the 2009 version. I already searched previously >> asked questions for my problem, but did not find an answer. >>> I want to pre-select e.g. Chip-Seq data from the genome browser >>> before >> sending it to Galaxy so that the file is not too big. To that end I >> upload my regions of interest as a custom track. >>> When using the 2006 version I go to "intersection", choose "custom >> track" and "All XY records that have an overlap with user track" >>> This works perfectly, I only retrieve the data for the regions of >> interest. >>> In the 2009 version I basically attempt to do the same thing >> (intersection has more options here, but the default is the same + I >> also tried the other possibilities, just to be sure), but it does not >> seem to work, judging on my output. >>> I wonder what goes wrong here - maybe you can help me? > >>> All the best - Janina >>> >>> >>> _______________________________________________ >>> Genome maillist - [email protected] >>> https://lists.soe.ucsc.edu/mailman/listinfo/genome >> _______________________________________________ >> Genome maillist - [email protected] >> https://lists.soe.ucsc.edu/mailman/listinfo/genome > _______________________________________________ > Genome maillist - [email protected] > https://lists.soe.ucsc.edu/mailman/listinfo/genome _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
