Friday, January 31, 2014

v32: shared metadata

From version 32, you are now able to create a custom metadata vocabulary and share/reuse it among your data hubs (instead of having to keep multiple copies over many hubs).

As an example, we put following JSON contents into a text file to define a metadata vocabulary:

{
vocabulary:{
    "epigenetic mark":{
        "dna methylation":["aa","bb","cc"],
        "histone mark":["dd","ee"],
        },
    "rna-seq":{
        "mRNA":["aa","ee"],
        "small RNA":["dd","bb"],
        },
},
tag:"My shared metadata",
}

To make it a shared metadata, put the file on a web server. Now the file is addressable by its URL:

http://vizhub.wustl.edu/hubSample/hg19/t/metadata


To share this metadata with a JSON data hub, add following JSON content into the hub:


{
type:"metadata",
vocabulary_file_url:'http://vizhub.wustl.edu/hubSample/hg19/t/metadata',
show:["epigenetic mark","rna-seq"],
},

The hub file can be found at this URL: http://vizhub.wustl.edu/hubSample/hg19/t/hub1withsharedMD



The shared metadata vocabulary will be in use.

Another hub which also shares this metadata vocabulary is here: http://vizhub.wustl.edu/hubSample/hg19/t/hub2withsharedMD

You can see how the shared metadata works by load this hub following the first one.


When a shared metadata is loaded, you can see the vocabulary by right-clicking term names on top of metadata color map, and choose "Add metadata terms" option:



The custom metadata is indicated by a tag name and orange background color:


Click the tag to get more information about this vocabulary:


Thursday, January 16, 2014

v31 (3 of 3): File upload

This function allows you to select text files from your computer and upload them to the browser server for browsing and analysis. Currently we provide limited supports on formats and complex data (BED, GFF, BEDGRAPH, WIGGLE etc.) but we plan to make steady and significant improvements to the File Upload function in near future.

This function is implemented in HTML5, and requires a modern web browser such as Google Chrome and Firefox.

We rely on the client-side to parse and process the file, and the amount of data that can be handled is determined by the user's computer capacity. So at this stage, avoid loading large files (e.g. >300MB) as it will consume a lot of RAM on your computer.

Go to "Apps" and "File upload" to see the interface:



Click the button to upload a file (or multiple of them).


Each uploaded file is shown as an entry. Click "Setup" to configure that file for upload:


This shows the configuration of a line-oriented BED file, with just 3 fields.

Check the radio button to choose appropriate file format, and either make a track out of this file, or make a gene set.

Currently for BED and GFF format data, there's only limited support. We are planning to make full support on these formats, as well as supporting new formats such as NarrowPeak.

v31 (2 of 3): Sessions

The session function has been on the browser for a long time, and it was completely redesigned to better suit its purpose, to make it robust and portable.

Before this point the sessions were stored in a mysql database on the browser server. We now switch to the file-based storage method, and creates text files to store session data.

And the data format is JSON (the browser loves that). And we enable you to download this file, keep it on your own computer, email it to a colleague, or even go to your own browser site and revive it (given all tracks are accessible).


Simply go to "Apps" > "Session" to open the session panel below and try it yourself:




Following shows how the saved session entries might look like. By clicking the "Download" button you can download this session as a JSON text file. To recover it click the "Upload" button on top.


v31 (1 of 3): Improved user interface

In version 31, we introduce a few changes to the user interface in the hope of improving user experience.

The greetings page, with a catalog of genomes:





Once you've chosen a genome, a few options appears in anticipation of your next move. You can add custom tracks, load public hub (if available), or skip these and go to browser:



A concise toolbar, now only with 3 buttons. Click the "Tracks" button to see a menu which allows you to access and manage all tracks in one place. Since we haven't got any tracks loaded, the menu below appears blank:



When we have a public data hub loaded such as Roadmap Epigenomics data set, we can see vital statistics in this menu, such as total track number, and those on show. Click the number area to open the facet panel (or track table) for track selection: