CyVerse logo

Home_Icon Learning Center Home

Building a Workflow

A workflow is a series of Apps chained together to run in sequence as a batch operation. On SciApps, workflows are generated from the completed analyses. Here we will show how to build a two-step workflow with SNAP and MAKER for a second round annotation.


Example Data

Input Description Example
Assembled genome A scaled-down genome that is comprised of the first 300kb of three chromosomes of rice test_genome.fasta
Annotated gene models MAKER output in GFF3 format (gzipped) maker_out.gff.gz
SNAP HMM estimation HMM parameters estimated with SNAP snap_out.hmm

Apps:

App name Version Description App link Notes/other links
MAKER 0.0.1 A portable and easily configurable genome annotation pipeline MAKER-0.0.1 MAKER documentation
SNAP 0.0.1 Semi-HMM-based Nucleic Acid Parser SNAP-0.0.1 SNAP documentation

Step 1: Running MAKER with SNAP output

This step will show you how to use SNAP output with MAKER to do a second around annotation.

  1. Login to SciApps at https://www.SciApps.org.

  2. Run SNAP as in the last section if you haven’t.

    Tip

    Jobs will disappear from the History panel when the browser gets refreshed. To load jobs back to the History panel, click ‘Workflow’ then ‘My jobs’ to select them.

    myjobs_window

  3. Click Annotation category (left panel) or search for MAKER, then click MAKER-2.32 to load the app form.

  4. Click SNAP-0.0.1 in the History panel to expand its outputs, then drag and drop snap_out.hmm into the SNAP HMM file field.

    build_workflow2

  5. Under “Genome sequence file” click Browse DataStore, then navigate the example (sciapps > example > maker); select test_genome.fasta and click ‘Select and Close’.

  6. Under “Maker annotations” click Browse DataStore, then navigate the example data (sciapps > example > maker); select maker_out.gff.gz and click ‘Select and Close’.

  7. Leave others as defaults, then click the “Submit Job” button.

  8. Once COMPLETED, click the ‘Visualization’ icon (next to the MAKER-2.32 job in the History panel), select jbrowse_out.view.tgz and then click ‘Visualize’ to visualize the annotation results in a genome browser.

    jbrowse_out

    Note

    Select the tracks in the left panel to visualize. For new analysis, it may take a while for visualization to be ready since the results need to be copied from CyVerse Data Store to the SciApps visualization server.


Step 2: Creating a Workflow

This step describes how to build a two-step workflow with previously completed MAKER and SNAP jobs.

  1. Check the checkboxes for step 1 (SNAP), and step 2 (MAKER) in the History panel, then click the ‘build a workflow’ link to load the Workflow building page.

    build_workflow3

    Tip

    History panel Checkboxes and the workflow building page are interactive. Use the ‘Select All’ or ‘Reset’ button to simplify the selection process.

  2. Click the ‘Build Workflow’ button to build a workflow with selected jobs. Once built, a workflow diagram will be open for visual inspection.

    annotation_workflow2

    Tip

    Click the ‘Top Down’ button to draw the workflow diagram vertically.

  3. On the ‘Workflow Diagram’, you can save the workflow. Your saved workflows will appear in ‘My workflows’ (under the ‘Workflow’ menu from top navigation panel).

    Tip

    While saving a workflow, you can rename it and/or add a short description.


Step 3: Running a Workflow

This step will show you how to run a workflow.

  1. Navigate to ‘Workflow’, then ‘My workflows’, to load the workflow you created and saved (in Step 2).

    myworkflows_window

    Note

    Five operations are supported for a selected workflow:

    • ‘Relaunch’: Display app forms (main panel) and load job histories (right panel)
    • ‘Visualize’: Display workflow diagram and load job histories to right panel
    • ‘Load’: Load job histories to right panel
    • ‘Share’: Get a direct link to the workflow for sharing workflow_URL
    • ‘Delete’: Delete the workflow

    Tip

    Double click on workflow name or description to make modifications.

  2. Alternatively, you can load the app forms and job histories directly if you have the direct link for a workflow. Scroll down the app forms in the main panel to either submit a new run or view the workflow diagram.

  3. Scroll down the main panel, then click Submit Workflow. You will be asked to confirm and prompted to check the job status in the right panel. Then a live workflow diagram will be displayed with real-time analysis status updates.

    Note

    running_workflow

    The colour of the app node will change when the status of the analysis changes:

    • ‘Yellow’: Pending
    • ‘Blue’: Running
    • ‘Green’: Completed
    • ‘Red’: Failed

    Tip

    A running workflow can also be saved then selected (in ‘My workflows’) to visualize its running status later on.


Summary

Using the apps SNAP and MAKER as examples, you have gotten an overview of how to use SciApps - from accessing data in CyVerse Data Store to launching jobs, building workflows, importing workflows, running workflows, and visualizing results.

More help and additional information

Search for an answer:
CyVerse Learning Center or CyVerse Wiki
Post your question to the user forum:
Ask CyVerse

Fix or improve this documentation: