Building Land User Guide
From Array Suite Wiki
OmicSoft Lands function as a central data hub to store and explore -omic data from public or internal sources. Lands built by OmicSoft are organized by disease focus into Collections, and each Land within a collection serves as a repository for thousands of -Omic samples. OncoLand, DiseaseLand, and GeneticsLand Collections are continually updated with new land content. Please see Introduction to Land Content for an overview of Lands that OmicSoft has curated and built for its customers.
Our customers can register their own sample data into the Land framework in order to facilitate internal discovery. Internal Lands can be merged with OmicSoft Lands by Creating Virtual Lands. This User Guide is designed to help our customers find the wiki documentation to assist in building an internal land, keeping with Best Practices that have been acquired by our Land Development Teams.
Please email email@example.com if you have questions not covered in this User Guide.
Land layers are physically stored on the ArrayServer in the directories that are specified in the ArrayServer configuration file [ArrayServer.cfg]]. Admin users can using the ArrayStudio GUI.
The output of the Create Land function is to create:
- the land folder will be created on the ArrayServer
- the land configuration file (LandName.cfg) will be generated with ArrayLand Configuration Options (saved as LandDirectory/LandName/LandName.cfg).
Admin users can update configurations to the land using Tools | Refresh Land.
Publish -Omic Data to Land
By default the Land will be added to the Internal Land Collections.
Open the Land and click Tools | Publish To Land. The related documentation below will provide details on how to create the Input Files.
Related Articles for storing measurement data for individual samples
Measurement data is published to the Land as ArrayLand Vector Files Publish ALV.
Related Articles for storing comparisons between two groups of samples
Comparisons are published as TLV.
- Building ComparisonLand From Raw Data
- Building ComparisonLand From External Inference Reports
Adding Metadata to Land
Omic data is anchored in the land to Sample Metadata. If a sample is published to the Land, the omic data will not be visible if the sample ID is not linked in the metadata.
Click Manage | Samples | Manage Land Sample Metadata:
There are a number of wiki pages that document 'how to's' for Sample Metadata. Below are links to the wiki pages to find answers to some of your questions:
- What are accepted input file formats? Manage Land Sample Metadata
- Where can I find an example of a Metadata Table? Sample MetaData File
- What are the Restricted Characters? See the entry on Row Names and Column Names on the Sample MetaData File page.
- In Lands built by OmicSoft, metadata is organized in Hierarchical trees. How do I organize the variables to appear as in the image below? See the section on Hierarchical Display on the the Sample MetaData File page.
- How do I organize my metadata file to combine multiple data types for the same SampleID? Use [IntegrationLevel]] and see the section on Integrating/combining datatypes for a sample on the Building Land From Raw Data page.
- Is it possible to add useful tool tip text? Can I search the land for an Alias of my term (such as Gender/Sex)? Yes, please see information on adding Variable Metadata.
- I know I what access users have in the Land using Manage User Access. But, can I make certain data within Projects or columns in my Metadata Restricted to certain users? Yes, although these data privacy features are only 'turned on' for GeneticsLands. Please see check out how to manage project-level access controls [] and 'IsPrivate' in Variable Metadata to hide columns/make restricted.
Helpful Hints for Adding/Updating Sample Metadata:
Take note of the number of samples (rows) in the metadata table. When Adding New Metadata, the new number of samples should be the sum of the prior number and the number of new samples being registered (0 + 870 = 870 in this case). If it is less, then one or more of the new sample IDs has conflicted with an existing sample ID, which means that the old metadata were over-written for the conflicting sample(s).
If sample metadata were overwritten, the simplest solution is to map the new samples to new unique IDs. First, restore the appropriate metadata for the existing samples, as necessary (load from the metadata text files you had previously used). Then, re-register the new samples with new IDs.