Getting started

Jobs

Every tool run on ProteinIQ creates a job. Jobs track your inputs, settings, and results so you can revisit and share your work.

Every time you run a tool on ProteinIQ, a job is created to track the request, process your data, and store results. You can view jobs for a single tool from that tool's page, or browse your entire team's job history across all tools at /app/jobs.

What is a job

A job represents a single execution of a tool with specific inputs and settings. Each job includes:

  • Input data: The sequences, structures, or molecules you provided
  • Settings: Tool-specific parameters you configured
  • Status: Current processing state
  • Progress: Percentage completion while the job runs
  • Results: Output data when processing finishes
  • Metadata: Job name, creation time, submitter, and credits used

Job lifecycle

Jobs move through several states as they are processed:

  • Pending: Created and waiting to start. Common for ML models that require GPU resources.
  • Queued: Waiting in a batch queue for dispatch. Appears when running batch jobs that exceed available concurrency.
  • Processing: The tool is actively running on your data. Simple conversions finish in seconds. ML predictions typically take 1 to 5 minutes.
  • Completed: Processing finished successfully. Results are ready to view and download.
  • Failed: Processing encountered an error. The error message describes what went wrong.
  • Cancelled: You cancelled the job before it finished.
  • Timeout: The job exceeded its execution time limit.
  • Retry: The system is re-running the job after a transient failure.

Most jobs you see will be Pending, Processing, Completed, or Failed. The other states appear in specific situations: Queued during batch runs, Cancelled when you stop a job, Timeout for oversized inputs, and Retry during automatic recovery.

Important

Credits are deducted when a job starts processing, not when it completes. Jobs that fail due to platform errors are refunded automatically.

Viewing jobs

You can view jobs in two places: the global jobs page and per-tool history.

Global jobs page

Navigate to /app/jobs to see all jobs your team has submitted across every tool. The list shows each job's name, status, duration, tool, and submitter. This is the best place to find a job when you do not remember which tool you used.

Per-tool history

From any tool page, click History in the toolbar to see only that tool's jobs. Click any job to load its results back into the tool view.

Job details

Opening a job shows the complete input data, all settings used, full results, credits consumed, execution time, and submitter. Jobs that are still processing update in real time without refreshing the page.

Managing jobs

Use the job menu (three dots) on any job to rename, delete, retry, duplicate, or export it.

Rename a job

  1. Click the job menu (three dots)
  2. Select Rename
  3. Enter a descriptive name
  4. Click Save

Use names like "Aspirin docking run 3" or "1ABC ESMFold 500-residue" to find experiments later without guessing.

Delete a job

  1. Click the job menu (three dots)
  2. Select Delete
  3. Confirm deletion

Warning

Deleting a job permanently removes all associated inputs, results, and metadata. Download any important files before deleting.

Retry a failed job

Failed jobs can be retried. Click the job menu and select Retry to create a new job with the same inputs and settings. Credits are charged for the new job as usual.

Duplicate a job

Completed jobs can be duplicated. Click the job menu and select Duplicate to create a copy. This is useful when you want to re-run an analysis with small changes to the settings.

Export results

Completed jobs with output data can be exported. Click the job menu, open the Download submenu, and choose CSV or JSON. CSV exports build columns dynamically from the output fields. JSON exports wrap results in an envelope with the job name, tool, settings, and timestamp.

Copy link and open in new tab

  • Copy link: Copies the job URL to your clipboard. The URL includes a ?jobId= parameter that loads that specific job.
  • Open in new tab: Opens the job in a separate browser tab.

Sharing jobs

Jobs can be shared with other people through the share dialog.

  1. Click the job menu (three dots)
  2. Select Share
  3. Choose a visibility level
  4. Click Save

Visibility levels

  • Private: Only you can view this job. This is the default.
  • Public: Anyone with the link can view this job.
  • Logged-in users: Any ProteinIQ user can view this job.
  • Invited only: Only people you invite by email can view this job.

When set to Invited only, enter email addresses and click Add. Users who already have a ProteinIQ account see the job immediately. Users who have not registered yet show a Pending badge until they create an account.

Note

Sharing controls who can view a job. Shared users can see inputs, settings, and results, but cannot edit or re-run your job.

Batch runs

Batch runs let you submit many inputs at once against a single tool. This is available for AutoDock Vina, Boltz-2, and AlphaFold-2.

Navigate to /app/batch, select a tool, configure your inputs, and submit. AutoDock Vina batch runs support up to 10 ligands in batch mode and offer three docking modes: single (one ligand per receptor), simultaneous (co-dock ligands together), or batch (dock each ligand independently). Boltz-2 and AlphaFold-2 use a list input: one complex or sequence per line.

Batch runs have a dedicated list page at /app/batch and a detail page showing progress for each child job. Child jobs do not appear in the main job history. Credits are calculated upfront and shown before submission. Failed child jobs are refunded automatically.

Credits and costs

Each job costs credits based on the tool type, input size, and compute resources required. The exact credit cost is shown on the Run button before you submit. Hovering over the button shows a breakdown of the estimate and your remaining balance.

Format converters and simple utilities cost 0 credits. Compute-heavy tools cost 3 to 250 credits depending on the tool and input size.

For a full breakdown of costs by tool category, see Credits.

Troubleshooting

If a job isn't behaving as expected, these are the most common causes.

  • Job stuck in pending: High server load or waiting for GPU resources is the most common cause. Wait a few minutes and check the status page for outages. Contact support if the job has been pending for more than 10 minutes.
  • Job stuck in queued: Batch jobs queue when concurrency limits are reached. They dispatch automatically as processing slots open.
  • Job failed immediately: Usually caused by invalid input format, missing required data, or a file exceeding size limits. Check the error message for details.
  • Job timed out: The input was too large or the tool hit its execution limit. Try reducing input size or splitting into smaller jobs.
  • Can't find an old job: Check the global jobs page at /app/jobs, which shows all tools. Per-tool history only shows that tool's jobs.
  • Results missing after completion: Refresh the page. If results are still absent after the status shows completed, contact support.

Next steps

  • Files: Browse, preview, and download files across all your jobs in one place
  • Credits: How credit costs are calculated and what tools cost
  • Quickstart: A complete walkthrough of running your first tool