Thanks everyone for the discussion! Here’s a quick recap of what we agreed on. Priorities, in order: 1. Get familiar with the pipeline: Iraaj will start by reading the ⁠HGTseq publication to understand the scientific rationale and design choices before diving into the code. 2. Nextflow compatibility: our first technical priority is to identify and address compatibility issues with recent Nextflow versions. The goal is to make only the necessary (strict) syntax/configuration changes, without altering the scientific workflow. 3. Module updates: BWA, SAMtools and MultiQC have already been updated in the dev branch, but haven’t been merged because we still need proper benchmarking. Before moving forward, we need to verify that the updates don’t change the results, ideally through byte-by-byte comparisons where applicable. 4. nf-core template updates: lowest priority (sorry 😅). These updates often concern development tooling rather than pipeline functionality, while potentially introducing extensive merge conflicts and manual work. We’ve already raised this issue with the nf-core community, so I’d rather focus our efforts on updates that bring actual benefits or resolve concrete problems. AI coding agents We’re absolutely happy to use AI coding agents, including for large-scale development work! The important thing is to first brainstorm and agree on a detailed, human-designed plan, with appropriate checks and validation steps in place. Once that’s clear, agents can be extremely useful for implementation. We just need to make sure we understand and verify the changes they introduce, rather than letting them drive the development decisions. Next steps Iraaj will read the publication, look into potential Nextflow compatibility issues and share his findings here on Slack. We’ll then discuss the proposed changes and agree on a plan before starting implementation. No need to schedule another meeting just yet. We can decide whether a Zoom call is useful once we have the initial assessment. The overall idea is to keep the pipeline scientifically reliable and up to date, without spending time on unnecessary changes.