Skip to content

Your First Workflow

This guide walks you through running your first bioinformatics pipeline on Flow.

Prerequisites

Before starting, ensure you have:

  1. Logged into the Admin Console or CLI
  2. Selected an organization and project
  3. Cloud credentials configured for your project

The CLI provides the fastest way to run workflows.

Step 1: Setup

# Install if not already done
pip install iseq-flow

# Login and select project
flow login --token YOUR_PAT_TOKEN
flow config select-project

Step 2: View Available Pipelines

flow pipelines list

Output:

SLUG                      NAME                      MODE                 ACTIVE
--------------------------------------------------------------------------------
hereditary-mock           Hereditary Mock Pipeline  direct_wdl           Yes
wdl-minimal               WDL Minimal Pipeline      direct_wdl           Yes
nextflow-minimal          nextflow-minimal          container_entrypoint Yes
nfcore-demo               nf-core/demo              direct_nextflow      Yes

Step 3: Get Pipeline Details

flow pipelines info hereditary-mock

Output:

Pipeline: Hereditary Mock Pipeline
  Slug: hereditary-mock
  Type: direct_wdl
  Container: broadinstitute/cromwell:86
  Tags: wdl, hereditary, mock, cromwell

  Description: Mock hereditary analysis pipeline for WDL/Cromwell testing.

  Required inputs: case_id, child_fastq

Step 4: Submit a Run

File paths are relative to your project bucket:

flow runs submit --pipeline hereditary-mock \
  -P case_id=test-case-001 \
  -P child_fastq=data/R1.fastq.gz \
  -P child_fastq=data/R2.fastq.gz \
  --watch

The --watch flag monitors progress until completion:

Run submitted successfully!
  ID: abc123-def456-...
  Name: hereditary-mock-20260113-120000
  Status: queued

Watching run status (Ctrl+C to stop)...
[12:00:15] Status: queued
[12:01:00] Status: running
           Started: 2026-01-13T12:00:55Z
[12:05:00] Status: succeeded

Run completed successfully!
Output: gs://your-bucket/outputs/hereditary-mock-20260113-120000/

Step 5: Check Run Status

# List recent runs
flow runs list

# Get details of a specific run
flow runs status RUN_ID

Step 6: Download Results

# List output files
flow files ls outputs/

# Download specific files
flow files download outputs/report.html
flow files download outputs/variants.vcf

Using the Web Interface

Step 1: Navigate to Compute

Go to the Compute service from the navigation menu.

Compute Dashboard

Step 2: View Available Pipelines

Click on Pipelines to see available workflows:

Pipelines List

Step 3: Launch a Run

  1. Click on a pipeline to view its details
  2. Click Run Pipeline
  3. Configure run parameters:
  4. Name: A descriptive name for this run
  5. Profile: Use docker,test for testing

  6. Container: Leave as default

  7. Click Submit

Step 4: Monitor Progress

Navigate to Runs to monitor your workflow:

Runs List

Run statuses:

Status Description
PENDING Run submitted, waiting to start
QUEUED Queued on GCP Batch
RUNNING Workflow executing
SUCCEEDED Completed successfully
FAILED Completed with errors

Step 5: View Results

Once complete, click on the run to view:

  • Execution logs
  • Output file locations
  • Timing and resource usage

Clinical Workflow Example

For clinical labs, associate runs with orders:

# 1. Create a clinical order
flow orders create -n "Patient Case 001" \
  --accession "ACC-2026-001" \
  --priority urgent \
  --indication "Suspected hereditary cancer"

# Order created with ID: order-123...

# 2. Submit run linked to order (paths relative to bucket)
flow runs submit --pipeline hereditary-mock \
  --order-id order-123... \
  -P case_id=patient-001 \
  -P child_fastq=data/R1.fastq.gz \
  -P child_fastq=data/R2.fastq.gz \
  --watch

# 3. Transition order status
flow orders transition order-123... analyzing

# 4. List runs for order
flow runs list --order-id order-123...

Next Steps