Hello Gabriel, You are correct, the statistics are generated for a complete alignment result. For transcripts, this would include all "exon" blocks if aligned to genomic reference sequence.
The exact format is described here, along with many other details that may be helpful to you: http://genome.ucsc.edu/FAQ/FAQblat http://genome.ucsc.edu/goldenPath/help/blatSpec.html -> File Formats -> .psl files The recommended option for gathering per-block statistics as you defined them is: -> create the alignment by running a BLAT and parse through the results using your own tools However, there may be other methods to experiment with, one example is: -> break up the transcript into block-segments (using the original alignment) and generate a 1-1 alignment versus the known target. This option could have complications if blocks are short, were merged during the original processing, or the run used specialized parameters (the RefGene track uses the BLAT parameter "-fine" to capture terminal exons). We hope this helps you to develop a solution, Jennifer ------------------------------------------------ Jennifer Jackson UCSC Genome Bioinformatics Group ----- "Gabriel Renaud" <[email protected]> wrote: > From: "Gabriel Renaud" <[email protected]> > To: [email protected] > Cc: "Fabio Passetti" <[email protected]> > Sent: Monday, January 11, 2010 6:31:37 AM GMT -08:00 US/Canada Pacific > Subject: [Genome] per exon identity scores for RefGene > > Hi, > A quick question. You have the number of mismatches on a per > transcript basis in RefSeqAli for RefGenes correct ? Is there any way > to > get the number of gaps/mismatches that were found in the alignment on > a > per exon basis instead ? > > Thank you, > > -- > > Gabriel Renaud > Laboratório de Bioinformática e Biologia Computacional > Serviço de Pesquisa Clínica > Instituto Nacional de Câncer (INCA) > Rio de Janeiro - RJ > Tel.: +55-21-3233-1343 > > > _______________________________________________ > Genome maillist - [email protected] > https://lists.soe.ucsc.edu/mailman/listinfo/genome _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
