Hello, The result is proably due to a techincal way that the data is stored and compared.
Data in these tracks are stranded and stored per exon (the introns are inferred from these coordinates). When the intersection was calculated, only the exons were used to identify exclusions (not the entire footprint of the gene i.e. introns + exons). Since you want to include the entire footprint of the gene, you will want save the tracks in othe formats first in BED format (from psl or genePred). Then, open that BED custom track and again save as BED, but with limited columns (the first six columns if you need strand - which is sounds like you do). The minimum BED file is 3 columns of data. http://genome.ucsc.edu/FAQ/FAQformat Also, using a higher quality or more complete gene track might be intersting to compare/use. The UCSC Gene's track contains the RefSeq track plus data from other trusted sources. See the track description for sources and methods. Good luck with your project - we hope this helps a bit, Jennifer Jackson ------------------------------------------------ Jennifer Jackson UCSC Genome Bioinformatics Group ----- "Soojin Yi" <[email protected]> wrote: > From: "Soojin Yi" <[email protected]> > To: [email protected] > Sent: Monday, August 17, 2009 1:30:50 PM GMT -08:00 US/Canada Pacific > Subject: [Genome] identifying regions devoid of specific tracks > > Dear Genome, > > I am trying to identify regions of the human genome that do not > contain any RefSeq or mRNAs. To achieve this goal, I first generated > a > custom track: > > I performed an intersection (AND) of > complement of RefSeq genes before base-pair-wise intersection > complement Human mRNAs before base-pair-wise intersection. > > I reasoned that the resulting custom track should be free of RefSeq > genes and human mRNAs. However, this was not the case when I manually > checked some of the custom tracks generated by this function. > > For example, > chr1: 183,279,941-183,280,644 > > contains one RefSeq gene (NWMAT2) and two mRNAs. > > Any idea how I can generate the custom track I want? > > Thanks, > > Soojin Yi > _______________________________________________ > Genome maillist - [email protected] > https://lists.soe.ucsc.edu/mailman/listinfo/genome _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
