Hello, I will explain using the table browser - although the table affyExonTissues is better downloaded because of the size. The tissue info you should extract from the browser. Merging and extracting data will require you to build your own tools - there is not an automated method.
To get the data: 1) Ftp the table from http://hgdownload.cse.ucsc.edu/goldenPath/hg18/database/ the table name is saved as a text file, same name, with .txt.gz added 2) Download from Table brower, with this track/table selected the output type "microarray names". Save locally. The field affyExonTissues.expIds is a list of the tissues from the "microarray names" file that are included in each data line. The field affyExonTissues.expScores is a list of the expression score for the range. These are ordered together: the third tissue goes with the third score, etc. The "bin" field is just for internal indexing, you can ignore it (and should when interpretting the data format, which is a variation of BED format special to ENCODE). http://genome.ucsc.edu/FAQ/FAQformat#format1 http://genome.ucsc.edu/FAQ/FAQformat#format6.5 This link will provide details about the values asscociated with this format. http://genomewiki.ucsc.edu/index.php/Microarray_track To understand the values assigned, the track description provides the details in the Methods section. See the Links for more details about the Affy analysis programs used. Thanks! Jennifer ------------------------------------------------ Jennifer Jackson UCSC Genome Bioinformatics Group ----- "Emanuele Osimo" <[email protected]> wrote: > From: "Emanuele Osimo" <[email protected]> > To: "UCSC mailing list" <[email protected]> > Sent: Friday, December 18, 2009 7:50:02 AM GMT -08:00 US/Canada Pacific > Subject: [Genome] how to use tissue specific expression data in table > affyExonTissues > > Hello everyone, > I'm new to this mailing list, so please forgive my simple question. > I'm trying to use the tissue specific genome wide expression data from > table > affyExonTissues in UCSC. > I see in the description that > > > *RNA (from a commercial source) from 11 tissues were hybridized to > > Affymetrix Human Exon 1.0 ST arrays. For each tissue, 3 replicate > > experiments were performed for a total of 33 arrays. The raw > intensity > > signal from the arrays was normalized with a quantile normalization > method, > > then run through the PLIER algorithm. The normalized data were then > > converted to median-centered log-ratios, which are displayed as > green for > > negative log-ratios (below-median expression), and red for positive > > (above-median expression).* > > > > What I cannot understand is how to obtain a unique representative > expression > value for each tissue, to be compared to the values in other tissues, > if > possible. > I understand that in the columns *expIds* and *expScores* contain the > 33 > experiments ids and results for the 11 tissues. But which experiment > is for > which tissue? And why are some expression values negative? > And why do many probes have the same *bin* value? And what does the > *score *column > stand for? > > Thank you very much in advance > > > Emanuele Osimo > Dept of Science & Biomedical Technologies - Fondazione Filarete and > School > of Medicine, University of Milan, Milan, Italy > _______________________________________________ > Genome maillist - [email protected] > https://lists.soe.ucsc.edu/mailman/listinfo/genome _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
