All tutorials
WorkspaceReady

Runs & Candidates

/runs-candidates?tab=runs

TL;DR

One workspace for execution history and candidate library review.

Use it / Skip it

A researcher needs both what ran and what survived. This page keeps run operations and candidate browsing in the same place.

Use when

You need to check progress, reopen outputs, inspect candidates, or hand off a ranked set.

Don't use for

Do not use it for module-specific setup. Start the run from the relevant page first.

Inputs

Tab
Runs for execution history, Candidates for scored library review.
Filters
Status, target, score, or text filters depending on the tab.

Outputs

Run or job record
A persisted job with status, timestamps, inputs, and enough context to reopen it later.
Artifacts
CSV, JSON, PDB, or report files depending on the module.
Decision signal
A shortlist, score, verdict, or review note that supports the next wet-lab or design step.

Walkthrough

  1. 1. Open Runs when something is queued, running, failed, or needs provenance

    Open Runs when something is queued, running, failed, or needs provenance.

  2. 2. Open Candidates when the question is what to express, export, or compare

    Open Candidates when the question is what to express, export, or compare.

  3. 3. Use filters before interpreting top rows

    Use filters before interpreting top rows. Different modules write different metadata.

  4. 4. Reopen the source module for deep inspection when a row needs structural or run-specific context

    Reopen the source module for deep inspection when a row needs structural or run-specific context.

Under the hood

  • Aggregates persisted run and candidate data written by platform modules.
  • Specific artifact availability depends on the module that created the record.

Worked example

Find last week's BCMA shortlist
Filter runs/candidates by BCMA and date, reopen the semi-de-novo or Interface source job, and export the candidate subset for review.

Pitfalls

  • A high score without provenance is weak evidence; always reopen the source job before handoff.
  • Old records may have fewer metrics than newer jobs after schema additions.

See also