I am very very new to Galaxy and bioinformatics and have been trying to navigate my way through this - I have multiple metagenomic shotgun libraries from multiple time points of honeybee gut samples. I imported them all, went through QC and filtering, and then I grouped them as a collection with paired forward and reverse reads. Now when I go to look at those files they come up totally blank, and if I go to run any analyses on them it says the dataset has been deleted (I didn't delete it). It is still showing it as taking up memory for me, just - nothing is there. Does anyone know how I can get this back or why it is happening? When I got to "undelete" it - it says there is an error. If I click on the [i] info icon while browsing histories, for individual reads in the collection, I can get them to open in a new window...so the data is clearly there somewhere! Thank you so much for any suggestions! It took me days to upload all the libraries so I don't want to lose them!
In the history Microbiome_Apis, I do some problems with the display of some of your dataset collections when clicking on the nested content.
Not all have this display problem. Some collections display the nested datasets with "blank" content (server issue, details below). Others display fine. Both represent collections that have queued jobs (not completed yet). The collections at the top history level are colored green and the jobs inside are colored gray. This history display/content is the expected behavior and all will eventually resolve to first to gray (queued), then yellow (executing), and later the jobs inside will run and all will be colored green or red (success or fail).
This "blank" collection content display problem is a server-side issue right now. Usegalaxy.org is under very heavy load and we are watching it closely.
Support FAQs: https://galaxyproject.org/support/
- Datasets and Histories https://galaxyproject.org/support/#datasets-and-histories
The Fastq collections seem fine. It does look like you chose to create some collections using the option to hide the original datasets. If you click on the collection, the individual datasets are listed and accessible (no tool jobs are involved -- creating a fastq data collection is just a data merge/rearrangement). If you want to access any or all of the individual datasets with tools (instead of the collection), use the toggle for "show/hide hidden" at the top of the history panel and unhide those you want to access directly as individual datasets.
I also see some deleted and some permanently deleted datasets in this same history, yet I don't see any problems with those with respect to the active collections. Permanently deleted data cannot be undeleted, only deleted data can be. If you delete or permanently delete the datasets that were included in a collection, the collection will still be in your history (as an artifact), but the datasets inside will have the deleted/permanently deleted status (since this is really the same data, just a rearrangement of the original). So, only delete/perm delete data if you no longer want to work with it.
Use "hide" if you want to unclutter the history when using collections. A collection represents merged links to the primary datasets and does not use up any additional quota space. If the quota calculation seems wrong, log out then back in again to reset it.
The last changes to this specific history were made about a day ago. The other two active histories do not contain any active collection datasets (eg: undeleted collection data).
If this problem really is at https://usegalaxy.org, could you specify which history and dataset number(s) are presenting with the delete/undelete issue? Usegalaxy.org has a different Galaxy release installed than most other Galaxy servers (the pre-release of 18.05, close to completion). Many collection enhancements have been added and the Galaxy application behavior will differ from prior releases. If there are still any bugs left, we want to learn about them and fix (if possible) before the official release.
And if you are working somewhere else, it is most likely that the other server has hit some problem we have addressed in the newer release. If you want to help test that, try to see if you can reproduce the same problem at usegalaxy.org. If you can, please report it back with the history/dataset name/numbers and we'll review.
Collections tutorial: https://galaxyproject.org/learn/
- Dataset collections - modern studies usually include many samples. Collection are designed to simplify complex, multi-sample analyses as shown in this tutorial. https://galaxyproject.org/tutorials/collections/
Thanks! Jen, Galaxy team