Home

Introduction

This wiki contains information on how to install, stage, configure and run the tglow-pipeline for analysing HCI imaging plates.

Primary pages:

1. Installation
2. Setup & staging data
3. Manifests & configuration
4. Configuring & running the pipeline
5. Understanding outputs
6. Analysing features in R
7. Guided example run

Additional resources: FAQ · Known issues · Options reference

Design notes

• The Nextflow pipeline runs in two main stages: stage (prepare data) and run_pipeline (processing).
• Most processes rely on Nextflow storeDir as a persistent cache between runs. If outputs exist in the store, tasks are not re-run; remove or re-name store entries to force re-execution.
• Parallelisation is done per-well to avoid many short tasks with high overhead (Conda activation, Python startup).
• The pipeline is field-aware and can handle missing fields across cycles or plates.
• The pipeline does not perform stitching. If you need stitched images, stitch before running the pipeline and disable flatfield estimation, which does not work on pre-stitched images.