Hello, Max You may have uncovered a bug in the Table Browser. The MAF file output (CDS FASTA...) is respecting the stop codon after 18 amino acids that is also visible in the graphic on the main Browser (chr17:62401176-62401178). Note that translation in right to left in the browser, as it matches the reverse strand.
However, when the refGene table is selected and the NM_000442 accession is specified in the list and the output is set to "sequence," the translation is 700+ AAs, as you report. This does match the record at NCBI for this accession: http://www.ncbi.nlm.nih.gov/nuccore/313760623?report=GenBank Because we do our own alignments of the RefSeqs, we do sometimes differ from the RefSeq record. We are looking not why the two paths through the Table Browser give inconsistent results, though it may take some time to sort it out. thanks for bringing this to our attention. --b0b kuhn ucsc genome bioinformatics group On 1/19/2011 3:54 PM, Maxwell Sanderford wrote: > Hi, > The outputs from the refGene table in the table browser are causing me some > confusion. I am using the table browser with the Feb. 2009 hg19 assembly, in > the refGene table in the hg19 database. What is confusing me is that the > sequence given by the sequence output option sometimes differs drastically > from the human sequence in the alignment given by the CDS fasta alignment > output option. > Example: Using the accession NM_000442 as a pasted identifier, the human > sequence in the in the CDS fasta alignment output is only 18 characters > long, whereas the sequence given by the sequence out option is 739 > characters long. > > Thanks, > ~Max > _______________________________________________ > Genome maillist - [email protected] > https://lists.soe.ucsc.edu/mailman/listinfo/genome _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
