Hi J, Thanks for your input, and thank you for the compliments on the Genome Browser!
-- Brooke Rhead UCSC Genome Bioinformatics Group On 3/24/12 1:44 PM, J Ireland wrote: > Hey Brooke, > > "On the other hand, the position format, and pretty much everything you > see in the graphical view on the main page > (http://genome.ucsc.edu/cgi-__bin/hgTracks > <http://genome.ucsc.edu/cgi-bin/hgTracks>) or when you click on an item > and go to its details page, is 1-based, fully-closed." > > Awesome. This is exactly the confirmation/explanation I was looking > for. Thanks! > > In case it's of use/interest, I have heard from some of the scientists > that I work with confusion over the hash marks as it's easy to forget if > the number goes to the base to the left or right - and only becomes > clear if you go to the beginning of the chrom (as you suggest). It was > actually one of these complaints that lead me to start this inquiry. In > my opinion, hashes above the bases might be clearer for a 1-based, > fully-closed convention (on the other hand, the hashes where they are > now are great for a 0-based, half-open convention!). With that said, > the UCSC Browser is still the first one I send folks to and definitely > the one I find most usable and intuitive for my work. Keep up the good > work. > > Thanks! > -J > > > > On Fri, Mar 23, 2012 at 5:32 PM, Brooke Rhead <[email protected] > <mailto:[email protected]>> wrote: > > Hi J, > > The 0-based, half-open convention is what we use for numbering in > tables. You see it when you deal with tables directly. On the > other hand, the position format, and pretty much everything you see > in the graphical view on the main page > (http://genome.ucsc.edu/cgi-__bin/hgTracks > <http://genome.ucsc.edu/cgi-bin/hgTracks>) or when you click on an > item and go to its details page, is 1-based, fully-closed. > > It might be helpful to go the the very beginning of a chromosome, > say, chr1:1-10, and see that the first base on the chromosome is > labeled "1". As you've noticed, the number labels are not centered > over the base labels (the A, T, C, G, or N) at that zoom level . . . > the numbers are next to the tick mark that marks the edge of the > base. It makes more sense that this is the convention when you zoom > out a bit, and see only every 5th or 10th base labeled. > > > -- > Brooke Rhead > UCSC Genome Bioinformatics Group > > > On 3/23/12 5:09 PM, J Ireland wrote: > > Awesome! Thanks, Brooke. That's perfect. > > Last question to put this to rest (this was my past of original > question) - when I zoom way in on the browser (say down to 5 or 10 > bases), am I correct that this is showing me your interval (not > position) format? That's how it appears to me. > > Thanks again, > -J > > > On Fri, Mar 23, 2012 at 4:55 PM, Brooke Rhead > <[email protected] <mailto:[email protected]> > <mailto:[email protected] <mailto:[email protected]>>> wrote: > > Hi J, > > You're right: rs12345 is a regular SNP and not an insertion. > In the > snp135 table, it is listed as: > > chr22 25855458 25855459 > > which should be interpreted as [25855458-25855459), 0-based. > > In the display, on the other hand, the location is > represented in > our "position" format: > > chr22:25855459-25855459 > > which should be interpreted as [25855459-25855459], 1-based. > > The coordinate transforms page that Hiram pointed out is > really helpful: > http://genomewiki.ucsc.edu/____index.php/Coordinate_____Transforms > <http://genomewiki.ucsc.edu/__index.php/Coordinate___Transforms> > > <http://genomewiki.ucsc.edu/__index.php/Coordinate___Transforms > <http://genomewiki.ucsc.edu/index.php/Coordinate_Transforms>> > > Sorry for the confusion! > > -- > Brooke Rhead > UCSC Genome Bioinformatics Group > > > > On 3/23/12 1:52 PM, J Ireland wrote: > > Hey Hiram, > > Sorry - am I missing something? rs12345 looks like a simple > bi-allelic SNP > (not an indel) which should have a length of 1. It's > ref allele > matches > the ref genome. The annotation page also says it has > length 1. > I don't > think this is the perfect case you were looking for.... > > > http://genome.ucsc.edu/cgi-____bin/hgc?hgsid=248751473&o=____25855458&t=25855459&g=____snp135Common&i=rs12345 > > <http://genome.ucsc.edu/cgi-__bin/hgc?hgsid=248751473&o=__25855458&t=25855459&g=__snp135Common&i=rs12345> > > > <http://genome.ucsc.edu/cgi-__bin/hgc?hgsid=248751473&o=__25855458&t=25855459&g=__snp135Common&i=rs12345 > > <http://genome.ucsc.edu/cgi-bin/hgc?hgsid=248751473&o=25855458&t=25855459&g=snp135Common&i=rs12345>> > > Yep, I've seen that link. > > Thanks again. Sorry - I never dreamed this would turn > into such > a marathon > thread! > > -J > > > On Fri, Mar 23, 2012 at 1:34 PM, Hiram > Clawson<[email protected] <mailto:[email protected]> > <mailto:[email protected] <mailto:[email protected]>>> wrote: > > > Good Afternoon J: > > This is a perfect case to illuminate the interval > notation. > Note the interval this item is declared to be in > [25855459-25855459) > Note the size of this interval: = end - start = > 25855459 - > 25855459 = 0 > > The length of this item is ZERO ! It doesn't exist ! > It can not be found in this sequence. It is > somewhere between > actual bases: [25855458-25855459) and > [25855459-25855460) > but not in this reference sequence. > > So, in one sense, you could say that the interval: > [25855459-25855459) > is actually talking about the gap between two bases. > The genome > browser does not show gaps at all, there are only > bases and they > are directly next to each other with no intervening > space. > We don't > do gaps between bases. When you enter such a > position, we can't > display it, we do the next best thing and display > one of the > bases > next to this non-existent item. > > I'm sure you have already seen our discussion: > http://genomewiki.ucsc.edu/____index.php/Coordinate_____Transforms > <http://genomewiki.ucsc.edu/__index.php/Coordinate___Transforms> > > <http://genomewiki.ucsc.edu/__index.php/Coordinate___Transforms > <http://genomewiki.ucsc.edu/index.php/Coordinate_Transforms>> > > -Hiram > > ----- Original Message ----- > From: "J Ireland"<mr.james.ireland@__gm__ail.com > <http://gmail.com> > <mailto:mr.james.ireland@__gmail.com > <mailto:[email protected]>>> > To: "Hiram Clawson"<[email protected] > <mailto:[email protected]> > <mailto:[email protected] <mailto:[email protected]>>> > Cc: [email protected] <mailto:[email protected]> > <mailto:[email protected] <mailto:[email protected]>> > Sent: Thursday, March 22, 2012 11:24:00 PM > Subject: Re: [Genome] base vs gap numbering > > Hey Hiram, > > > I'm definitely a firm believer in the interval, > trust me - and I > definitely dig the UCSC data tables and the half-open, > zero-start intervals. > > > So, rolling with the "everything is an interval" and > looking > up rs12345... > > > If I zoom into the browser here: > > http://genome.ucsc.edu/cgi-____bin/hgTracks?db=hg19&position=____chr22%3A25855458-25855460 > > <http://genome.ucsc.edu/cgi-__bin/hgTracks?db=hg19&position=__chr22%3A25855458-25855460> > > > <http://genome.ucsc.edu/cgi-__bin/hgTracks?db=hg19&position=__chr22%3A25855458-25855460 > > <http://genome.ucsc.edu/cgi-bin/hgTracks?db=hg19&position=chr22%3A25855458-25855460>> > I see rs12345 appears to be at [25855458-25855459) > given the > hash marks. > This is a zero-based start and end. > > > Now, I click on rs12345 and see the position as > 25855459-25855459 in the > annotation page. Should I interpret this as > [25855459-25855459), with a > one-based start and zero-based end or as > [25855459-25855459] > a fully closed > interval with start and end both one-based? > > > I know you're straddling different count-starting > conventions between db > tables and the annotation pages. Maybe my > misunderstanding > is I thought the > browser was using the same convention as the annotation > pages, but it looks > like the browser is actually more inline with the db > tables... > > > Thanks for taking this trip with me down the rabbit > hole of > the obscure > and arcane ;) > -J > > ___________________________________________________ > Genome maillist - [email protected] > <mailto:[email protected]> <mailto:[email protected] > <mailto:[email protected]>> > https://lists.soe.ucsc.edu/____mailman/listinfo/genome > <https://lists.soe.ucsc.edu/__mailman/listinfo/genome> > <https://lists.soe.ucsc.edu/__mailman/listinfo/genome > <https://lists.soe.ucsc.edu/mailman/listinfo/genome>> > > > _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
