Hello Jen,
For this problem, you will need to reach the support or administrators for this public server. https://wiki.galaxyproject.org/PublicGalaxyServers#ODoSE
The authors of the primary publication are probably the best contacts (I couldn't locate any others). You can find these linked from the paper, which is open source, but I will also send them to you direct (to avoid posting to a public list without permission first).
Best,
Jen Galaxy team
On 12/31/13 9:18 AM, Jen wrote:
hello,
I've run across a possible bug in ODoSE that is preventing me from analyzing my data. I've scoured the various help documentation without luck, so I'm writing you.
I'm using the program via this site: http://www.odose.nl/
The history showing the steps below is here:
http://www.odose.nl/u/publ/h/unnamed-history
The problem is this: I'm unable to work with a set of genomes that is exclusively "user-generated" (i.e., external, and uploaded from my local drive).
To illustrate this problem with an example, I downloaded the .ffn files for four published genomes NCBI's ftp site. (This way, I'm sure my 'external' files are formatted correctly.)
I've tried to follow these instructions from http://www.odose.nl/ (Analyze Data tab) -
"User-generated whole-genome data not yet archived in GenBank can be uploaded to be analysed separately or in conjunction with published genomes. To do this, first create a new history: in the right ‘History’ pane go to Options > Create New. In the left ‘Tools’ pane go to Get Data > Upload File and select one or multiple .ffn files from your computer and click ‘Execute’..."
This worked fine.
"Go to Divergence > Upload genomes in the ‘Tools’ pane, press the ‘Add new genome’ button, specify a genome label and press ‘Execute’ for each genome to be included…"
I was unable to find a "Divergence" option in the Tools pane. So, I assumed this should be O'DoSE > Upload Genomes, in the Tools Pane. Using this function, I uploaded each genome and selected "Execute."
"Proceed to the Workflow menu to add listed genomes and/or to specify input settings…"
Perhaps this is my downfall, but I was unable to see how to add genomes in the Workflow menu.
At this point, I had uploaded four genomes, and attempted to select the first option in the analysis: "Extract & Translate Genes." I chose the option of selecting my external genomes, but the analysis failed, giving the error "History does not include a dataset of the required format/build" error. This error appears within the "Genbank Project ID's" box. In this case, I have no Genbank Project ID's, because all data are external genomes.
Interestingly:
The problem is solved by selecting at least one genome from the Complete Microbial Genomes table. Selecting one genome generates a Project ID list. I am then able to perform "Extract & Translate Genes" and all further functions, acting on the set of external genomes and the one selected (internal) genome. I've tried this several times, with different external genome files. I can successfully analyze dozens of external genomes, but _only if_ at least one 'internal' genome is selected as well.
I need to run an analysis on a set of genomes that is exclusively unpublished, so the data will be entirely "external" genomes. Including a published genome will throw off the comparisons we need to make.
Based on the ODoSE instructions above (that user-generated whole-genome data...can be uploaded to be analysed separately or in conjunction with published genomes), I sense I should be able to use exclusively 'external' genomes in an analysis. Do you have any suggestions on how I can do this? Any advice would be much appreciated.
thank you, Jen
(Apologies to the moderator for the earlier version of this message that was too large. Above, I've removed the attachments showing screen shots. I hope the problem is clear from the History link above.)
Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/
To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/