sample form creation, nglims, seq tracking
I'm using the nglims build via "hg clone http://www.bx.psu.edu/hg/galaxy galaxy_dist" as mentioned in the wiki, Apache proxy and Postgres DB enabled. Please correct me if any of my initial "goals" below are unattainable or if I am misinterpreting Galaxy's capabilities. I'm still attempting to learn the overall system capabilities of Galaxy and its integration between nglims samples+projects and the native Sample/Seq tracking components. Goals I'm trying to accomplish: (1) Create a fresh custom sample form in the Admin tab's "Manage Form Definitions" to use in the Lab tab's (nglims section) "next gen sample submission" process, and assign this custom sample form to be used as the "default" form when creating a new sample in nglim section's "next gen sample submission", (2) Allow multiple/various defined samples created with this custom sample submission form to be submitted in the Lab tab's "Submit samples as a project" as a project for tracking in the Admin tab section, (3) Allow "Find Samples" in the Admin tab to potentially query for nglims samples created with this sample form. Uncertainties: (1) Can samples created in the nglims section currently be tracked in the Admin tab's "Sample Tracking" area, or does the sample tracking functionality only allow tracking of the samples created with the native form definitions (within the Admin section, outside of the Lab tab "next gen sample submission")? There was a prior posting a while back that mentioning the following (verbatim): "You probably don't want to mix the "nglims" approach with Galaxy's native sample tracking functionality. While I've build the nglims part to use the same database tables and be fairly interoperable, it does take some different data representation choices, especially with regards to the sample/request relationship." Does this mean we're limited to creating samples outside of the nglims section if we want to track them within the native sequencing section? Secondly, will there be future support for tracking nglims-created samples with the native sample/seq tracking? (2) How to change which sample form is used in the Lab tab's "Define samples and services". In fact, I cannot determine what form is actually used which corresponds to the form which appears when one clicks the Lab tab's "next gen sample submission". Additional concern: When creating and/or editing a form in the Admin tab's "manage form definitions", I've created a sample submission form with multiple grids, some grids which have multiple fields beneath. When editing this form and assigning fields to the respective grids, each time I save and re-edit the form, all of the fields get reset to whatever is listed as the 1st grid in the "Select the grid layout to place this field" dropdown menu, rendering all of my assigned fields to be reset to the first grid. Adding or removing a grid with the submit button also does the same thing. Any thoughts? Galaxy appears to be a very promising product and we're looking forward to having the researchers and sequence team use this environment, once we finalize the configuration and push into production. I'm sure I'll have additional questions once I work through some of these current issues. My apologies for the extended length of this post and if I've somehow misinterpreted some of the functionalities and compatibilities. Thanks in advance for your feedback. Best Regards, Mike
Mike;
I'm using the nglims build via "hg clone http://www.bx.psu.edu/hg/galaxy galaxy_dist" as mentioned in the wiki, Apache proxy and Postgres DB enabled.
Please correct me if any of my initial "goals" below are unattainable or if I am misinterpreting Galaxy's capabilities. I'm still attempting to learn the overall system capabilities of Galaxy and its integration between nglims samples+projects and the native Sample/Seq tracking components.
Glad to have you taking a look at the LIMS functionality. Sorry for any confusion; let me try to clear up the different options in Galaxy: - The official functionality in Galaxy is referred to as Sample Tracking, with documentation here: http://main.g2.bx.psu.edu/u/rkchak/p/sts - nglims is an add-on that I'm developing; it's meant to work with the same underlying infrastructure as the sample tracking but is not an official Galaxy project. The documentation is here: https://bitbucket.org/galaxy/galaxy-central/wiki/LIMS/nglims and the fork that tracks galaxy-central, but with the nglims additions is here: https://bitbucket.org/chapmanb/galaxy-central
Uncertainties: (1) Can samples created in the nglims section currently be tracked in the Admin tab's "Sample Tracking" area, or does the sample tracking functionality only allow tracking of the samples created with the native form definitions (within the Admin section, outside of the Lab tab "next gen sample submission")? There was a prior posting a while back that mentioning the following (verbatim)
"You probably don't want to mix the "nglims" approach with Galaxy's native sample tracking functionality. While I've build the nglims part to use the same database tables and be fairly interoperable, it does take some different data representation choices, especially with regards to the sample/request relationship."
Does this mean we're limited to creating samples outside of the nglims section if we want to track them within the native sequencing section? Secondly, will there be future support for tracking nglims-created samples with the native sample/seq tracking?
Right, so the Admin page is for dealing with Sample Tracking only, not any of the nglims. The nglims is a separate interface which has all the things you've listed built in. The 'View projects' link allows you to search samples. The 'Queues' and 'Runs' pages allow you to assign samples to flowcells and manage them through the sequencing process. So if you'd like to go the nglims route, you don't need to use the Admin page at all.
(2) How to change which sample form is used in the Lab tab's "Define samples and services". In fact, I cannot determine what form is actually used which corresponds to the form which appears when one clicks the Lab tab's "next gen sample submission".
There is a configuration file in tool-data/nglims.yaml that defines the forms. You'll also see them in the 'Manage form definitions' page since the same database tables are used, but the configuration and associated install script (scripts/nglims/add_ng_defaults.py) automate the form creation.
When creating and/or editing a form in the Admin tab's "manage form definitions", I've created a sample submission form with multiple grids, some grids which have multiple fields beneath. When editing this form and assigning fields to the respective grids, each time I save and re-edit the form, all of the fields get reset to whatever is listed as the 1st grid in the "Select the grid layout to place this field" dropdown menu, rendering all of my assigned fields to be reset to the first grid. Adding or removing a grid with the submit button also does the same thing. Any thoughts?
The nglims does not use the grid layout, but should also not reset it. If you decide to stick with sample tracking, instead of nglims, rc might be able to help you with this. Sorry for any confusion; hope this clears things up. I'm also planning to get together a detailed overview of the nglims and automated backend processing we've put together, so hopefully more documentation will continue to make things clearer. Thanks, Brad
Hi Brad, Nice work on the new updates of nglims. We like that we can now build a flowcell. We know see a results form. How do we configure this form to show the results as well as the default plots we know see in the default layout? And also is it possible to add a data library to a project without using the native tracking request from the admin tab? We typically use one file per lane say the s_1_sequence.txt. Thanks in advance. Victor
Victor; Thanks much for playing with nglims. Glad things are working well for you with the interface and you are on to results processing.
Nice work on the new updates of nglims. We like that we can now build a flowcell. We know see a results form. How do we configure this form to show the results as well as the default plots we know see in the default layout?
The results are uploaded by the automated processing pipeline, which is available from: https://github.com/chapmanb/bcbb/tree/master/nextgen This generates fastq files as runs come off the sequencer, transfers them to the Galaxy machine, does alignments and other analysis steps like SNP calling, and then uploads the results to Galaxy. As part of that process, statistics on clusters and passing clusters are uploaded for display in the results and plots. This helps us monitor how things are going with the machine and diagnose issues. Happy to help with setting this up if you have any issues. The GitHub page has some basic documentation to get started.
And also is it possible to add a data library to a project without using the native tracking request from the admin tab? We typically use one file per lane say the s_1_sequence.txt.
What we do is have users report a lab group or affiliation when they create a Galaxy account, and then assign them permissions to a Data Library folder associated with that lab. When uploading, the fastq and other files go within a sub-folder of that library named by the date of the run. We get a lot of new users trying to submit requests who don't yet understand Data Libraries, so this workflow eases the learning curve for them. I'm finishing up a new entry page that supports assigning barcodes to samples which should be ready next week, and then hope to write up documentation so knowing the areas where you are running into issue is really helpful. Thanks much, Brad
Happy to help with setting this up if you have any issues. The GitHub page has some basic documentation to get started. Yes, of course. I'll be happy to test this out and looking forward to see it in action. I'll work on getting BCBB installed at out end and report any issues that I find. Pretty much that was the last thing we needed to start putting NGLIMS into a more
Thank you for your reply. I'm very impressed. production level. (connecting the results to a data library). I will be also happy to help with the documentation if you need me to. Please keep us posted on the new updates. Thanks again. Victor On Dec 18, 2010, at 1:51 PM, Brad Chapman wrote:
Victor; Thanks much for playing with nglims. Glad things are working well for you with the interface and you are on to results processing.
Nice work on the new updates of nglims. We like that we can now build a flowcell. We know see a results form. How do we configure this form to show the results as well as the default plots we know see in the default layout?
The results are uploaded by the automated processing pipeline, which is available from:
https://github.com/chapmanb/bcbb/tree/master/nextgen
This generates fastq files as runs come off the sequencer, transfers them to the Galaxy machine, does alignments and other analysis steps like SNP calling, and then uploads the results to Galaxy.
As part of that process, statistics on clusters and passing clusters are uploaded for display in the results and plots. This helps us monitor how things are going with the machine and diagnose issues.
Happy to help with setting this up if you have any issues. The GitHub page has some basic documentation to get started.
And also is it possible to add a data library to a project without using the native tracking request from the admin tab? We typically use one file per lane say the s_1_sequence.txt.
What we do is have users report a lab group or affiliation when they create a Galaxy account, and then assign them permissions to a Data Library folder associated with that lab. When uploading, the fastq and other files go within a sub-folder of that library named by the date of the run. We get a lot of new users trying to submit requests who don't yet understand Data Libraries, so this workflow eases the learning curve for them.
I'm finishing up a new entry page that supports assigning barcodes to samples which should be ready next week, and then hope to write up documentation so knowing the areas where you are running into issue is really helpful.
Thanks much, Brad
participants (3)
-
Brad Chapman -
Collins, Mike -
Victor Ruotti