Dear Martin, You are right. We don't support protein-level MAF alignments. Unfortunately, your suggestions would require a significant amount of work on our end, but we will pass them along to our developers. Thank you for your input.
Vanessa Kirkup Swing UCSC Genome Bioinformatics Group ----- Original Message ----- From: "Martin Frith" <[email protected]> To: "Angie Hinrichs" <[email protected]> Cc: [email protected] Sent: Wednesday, October 20, 2010 6:52:53 PM GMT -08:00 US/Canada Pacific Subject: Re: [Genome] custom track: translated psl strand bug Hi Angie and Galt, many thanks for your informative replies. > Did MAF not work for you because it is a nucleotide-level format, or > some other reason? It seems to me that MAF is a nice, general alignment format, which can handle protein-level alignments too. For example: track name="CDS homology" ##maf version=1 a score=150 s A4FUW6 94 44 + 221 SYMVSLEVGEEKHPVHLCKIMLFF-RSNPYFQNKVITKEYLVNIT s chrY 51817253 131 - 59373566 SSTLS*QVEEVKHPVHLCKIMLFF/SDAPYFQNKVITEEYLVNVT I tried uploading this as a custom track. First, I have to replace "chrY" with "hg19.chrY". Then it uploads successfully, but the display is messed up. I guess that your software does not currently allow for protein-level MAF alignments? I appreciate you are all busy, and indeed you are doing a wonderful job with the browser. It would be even more wonderful if it could handle protein-level MAF. (This format includes the query sequences, which solves one problem you mentioned.) Just my hope... Have a nice day, Martin On Wed, Oct 20, 2010 at 02:56:44PM -0700, Angie Hinrichs wrote: > Hi Martin, > > Regarding this part of your question: > > > > By the way, can I / should I use any format other than PSL for a > > > custom track of translated alignments? I tried MAF, but it didn't > > > work. It would be nice if I could click on a track item and see the > > > alignment. It would be nice if stop codons and frameshifts were > > > highlighted in red or something, is it possible to do that? > > We don't have another format specifically for translated alignments. Did MAF > not work for you because it is a nucleotide-level format, or some other > reason? If you do have nucleotide-level alignments, and can output the BAM > format (http://genome.ucsc.edu/goldenPath/help/bam.html), nucleotide-level > differences are highlighted in the same way as hgBlat untranslated results. > > It would be a non-trivial enhancement to highlight codon differences in PSL > custom tracks -- we would need to add a way for you to upload query sequences > associated with your alignments (and for nucleotide-level PSL, CDS start and > stop for each query), and we don't yet display translated differences even in > in hgBlat results. That would be very cool, but we all have pretty full > plates here already. > > Angie > > ----- "Galt Barber" <[email protected]> wrote: > > > From: "Galt Barber" <[email protected]> > > To: "Martin Frith" <[email protected]> > > Cc: [email protected] > > Sent: Tuesday, October 19, 2010 4:43:36 PM GMT -08:00 US/Canada Pacific > > Subject: Re: [Genome] custom track: translated psl strand bug > > > > Hi, Martin! > > > > This does seem to be a bug in the PSL customTrack loader. > > It should mark the strand as "-" instead of "+". > > We will look into it. > > > > Currently we convert PSL custom tracks to bed12. > > So it really can't show un-aligned parts. > > > > The BLAT track (hgBlat) does a much better job > > of showing more details about the alignment. > > > > Thank you for reporting the problem. > > > > -Galt > > > > 10/18/2010 7:05 PM, Martin Frith: > > > Dear genome folks, > > > > > > I would like to display a custom track of protein-versus-genome > > > alignments. I tried using PSL format, but there seems to be a bug > > > with "-" strands. > > > > > > For example, I blatted this sequence against hg19 with "output > > type" > > > set to hyperlink: > > > LSVADHAPQFAVPSDTATAPQAESVLRPLYMDVQATTPL > > > > > > The output looks like this: > > > SCORE START END QSIZE IDENTITY CHRO STRAND START END SPAN > > > 63 1 39 39 77.0% 20 +- 34286403 34286519 > > 117 > > > When I view it in the browser, the arrows go right-to-left, which > > is > > > correct. > > > > > > I repeated this blat with "output type" set to psl: > > > match mis- rep. N's Q gap Q gap T gap T gap > > strand ... > > > match match count bases count bases > > ... > > > 30 9 0 0 0 0 0 0 +- > > ... > > > I then pasted this PSL line into "add custom tracks". > > > When I view it in the browser, the arrows go left-to-right, which > > > seems wrong. > > > > > > > > > By the way, can I / should I use any format other than PSL for a > > > custom track of translated alignments? I tried MAF, but it didn't > > > work. It would be nice if I could click on a track item and see > > the > > > alignment. It would be nice if stop codons and frameshifts were > > > highlighted in red or something, is it possible to do that? > > > > > > Have a nice day, > > > Martin Frith > > > http://www.cbrc.jp/~martin/ > > > _______________________________________________ > > > 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
