Sbp 250

[vtnphan]

Pull Request

Summary

SBP-250 Replace Seqera Platform dataset uploads with direct S3 uploads for workflow samplesheets. This removes the dependency on Seqera datasets as an intermediary and simplifies the data flow: CSV samplesheets are now uploaded directly to S3, and the resulting S3 key is passed to workflow launchers.

Changes

• Removed seqera_dataset_id column from the WorkflowRun model and added an Alembic migration (drop_seqera_dataset_id_from_workflow_runs) to drop it from the DB
• Replaced Seqera dataset service (create_seqera_dataset, upload_dataset_to_seqera, upload_interaction_screening_dataset) with S3-backed equivalents (upload_csv_to_s3, upload_interaction_screening_csv_to_s3) in app/services/datasets.py
• Changed launch payload: WorkflowLaunchPayload.datasetId → s3InputKey; all three workflow executors (bindflow, wisps, proteinfold) now receive s3_input_key: str instead of dataset_id
• Updated response schemas: DatasetUploadResponse / InteractionScreeningDatasetUploadResponse → S3DatasetUploadResponse / InteractionScreeningS3UploadResponse
• Added RunInput tracking: on workflow launch, the samplesheet S3 object is recorded as a RunInput linked to the WorkflowRun, enabling a new presigned-URL endpoint to download the input samplesheet
• Updated schema diagram to reflect the removed column
• All tests updated across 10+ test files; coverage maintained at ≥90%

How to Test

1. Run the Alembic migration: uv run alembic upgrade head
2. Upload a samplesheet via the dataset upload endpoint and confirm an S3 key is returned (no Seqera dataset ID)
3. Launch a workflow with s3InputKey in the payload and confirm it succeeds
4. Verify old datasetId payloads are rejected with 422
5. Run the full test suite: uv run pytest --cov=app --cov-fail-under=90
6. Run linters: uv run ruff check app tests && uv run black --check app tests && uv run mypy app

Type of change

• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)

Breaking: WorkflowLaunchPayload.datasetId is replaced by s3InputKey. Frontend clients must update the launch payload field name. The seqera_dataset_id column is dropped from workflow_runs.

Checklist

• I have added tests that prove my fix is effective or that my feature works
• I have added or updated documentation where necessary
• I have run linting and unit tests locally
• The code follows the project's style guidelines