-
Notifications
You must be signed in to change notification settings - Fork 0
Home
In style of its eponym, palantir is a gateway browse other worlds, namely processed count data and transformations from experiments such as RNA-seq and Screens. Palantir is a deployable Maven 3 project compliant with Java EE6 on WildFly AS 10.x. It includes a persistence unit connecting to a MySQL datasource with processed and prepared NGS data utilizing palantir-processor and palantir-importer and transformations from experiments such as RNA-seq and Screens. The JSF front-end lets you conveniently browse the data and add metainformation such as sample annotations, screen candidate selection and comments, as well as grouping data into experiments.
Palantir was tested on Mozilla Firefox, Safari and Opera. Google Chrome is explicitly not supported (not for a lack of trying 👎 ).
NOTE: This is a very preliminary alpha release with only a limited number of features. The list will be updated with any additions as the implementation progresses.
All provided features are accessible via the left-hand sidebar menu.
The Samples section gives you an overview of all available samples with meta-information. This meta-information is currently fetched from the VBCF-Queue, so anything you enter there will be propagated to the database on demand.
On the bottom of the Sample-list you will find an edit Button. This allows you to edit meta-information of any of the samples. A click on the right-hand search button would erase all changes you made and replace them with the default information loaded from the VBCF-Queue from this sample.
Clicking on any of the samples directs you to the reference page. Here you will see for which references, count data is available.
The final page of the samples process is the genes page. This page contains count information as well as count transformations such as RPKM and TPM. The genes field let's you search for genes, all remaining columns can also be sorted on top. The search results and the entire table can be exported as csv, pdf and xls files by the corresponding export icons on top of the table / at the bottom, respectively.
NOTE: This is only a mockup implementation so far not leading to any persisted results. It is used to demonstrate a mechanism how grouping samples could work.
This page again displays all samples and let's you multi-select (holding STRG / COMMAND while clicking) samples you want to group into an experiment.
This page again displays all available count data with references used for the samples. You can again multi-select your desired count data, but only proceed if all of the selected references are identical.
This page summarizes the samples and count data with references you have selected. It will ask you to name the experiment and give a description before submitting the information. NOTE: Nothing will happen if you click submit except for an info message.
On the selection page, you will first have to select which reference you want to query. Once you have clicked a reference, all available genes will be loaded into the table below. Now you can select any gene you want and click it.
This page will now list for all samples counted on this reference the respective count data and transformations. Again the table can be exported as csv, pdf and xls files.