Hi, I'm have been using canfam3 from built-in index for TopHat and refflat downloaded from UCSC for Cufflink. Now I'm at Cuffmerge step howerver the reference genome selection doesn't offer canfam3. The only choice is canfam2. Can I still use it or should I download canfam3 from UCSC?. If I have to download, which file should I use: canfam3.fa.gz or canfam3.2bit? Thank you very much for your help.
Hi Jen, thank you very much for the reply. I wanted to refer to the reference genome. When choosing "yes" in use sequence data, a reference genome is required, but the pulldown menu has only one choice which is canfam2. Since I used canfam3 in tophat and cufflink steps, I think I have to stick with canfam3. I will appreciate very much if you can advise me how to obtain the canfam3 refence genome.
Best,
Hello,
Thanks for clarifying. I can confirm that there is a small bug in one of the configuration files. This impacts the description of the genome only. We are correcting the file, but meanwhile, if the input GTF dataset to CuffMerge has the database "canFam3" assigned, then the genome that is selected for the sequence option is actually canFam3 (although the description label states it is canFam2). The underlying dbkey associated with the data is canFam3 and is what would be used in the analysis and labeled in the output. No other similar errors have been found.
The other genome, canFam2 is intact without any description (or other) errors, if anyone is using GTF dataset with that database assigned.
Thanks again for reporting and helping to root out the core issue! Apologies for the inconvenience, Jen, Galaxy team
Hello again, We were able to make the description correction today and I just verified this on Main (http://usegalaxy.org). Thanks again for reporting the problem, Jen, Galaxy team
Hello,
The reference genome is not coming up or do you mean the reference annotation dataset? This is a GTF/GFF3 dataset. It is not provided within Galaxy and must be obtained from a source such as iGenomes (if available), UCSC, Ensembl, and others.
iGenomes is the best source due to specific attributes included in the annotation. I double check and we are in luck - canFam3 is supported.
http://support.illumina.com/sequencing/sequencing_software/igenome.html
Download the .tar file for the build = UCSC and genome canFam3. Uncompress the tarball, locate the "genes.gtf" dataset, upload just that to Galaxy, and double check the metadata assignments (datatype and database - use the "Edit Attributes" functions found by clicking on the pencil icon in the upper right corner for the dataset once loaded. You can set these upon upload, but it is always good to do the metadata double check to avoid problems later on.
Galaxy does not uncompress and load multi-file tar archives through the Upload tool. If you unsure about how to download the data from iGenomes, please see the instructions there. The commands needed to uncompress a tar archive can be found with a google search. This one is a good one to follow, with clear, specific commands listed for various OS systems and specific extensions. But the others that come up in the search are also good to review if you have issues.
http://www.cyberciti.biz/faq/linux-unix-bsd-extract-targz-file/
I perhaps added more detail here than you yourself require, but it may help others with a similar need. I will also add these instructions to my list of advanced topic to update/include during the help resource updates we are working on over the next month or so (these will be published to our wiki and here in Biostar for reference).
Hopefully this helps! Jen, Galaxy team
