---
title: Framework
tags: march-2021, nf-core, hackathon, talk
description: "Everything that's not a pipeline or a module."
---
{%hackmd theme-dark %}

# Hackathon March 2021
# Framework
The wheels of nf-core keep turning thanks to a body of code that we often loosely refer to as _"The Framework"_. This is a bit of a catch-all term that covers anything that isn't a pipeline.
The main projects are the `nf-core/tools` helper package and the website, though some lines are a little blurred between different repositories (eg. automation tools in pipelines propagated from the helper tools template).
# Group members
<!-- TODO: add yourself here! -->
- Phil Ewels
- Matthias Hörtenhuber
- Lukas Heumos
- Nathan Sheffield
- Gisela Gabernet
# nf-core/tools
- _[Monday]_ Collected lots of bugs from people after the last release
- Now in a [`v1.13.2` patch fix release milestone](https://github.com/nf-core/tools/milestone/11)
- _[Tuesday - Phil]_: Fix the template sync ([nf-core/tools#911](https://github.com/nf-core/tools/issues/911))
- _[Tuesday - Phil, Kevin]_: more bug fixes and release of v1.13.2
- _[Wednesday - Phil, Kevin]_: more bug fixes and prep for release ov v.13.3 :man-facepalming:
## DSL2 Modules
- _[Never happend - Phil]_: Write DSL2 module installation versioning ([nf-core/tools#868](https://github.com/nf-core/tools/issues/868))
# Website
- _[Mashehu]_: add database backend ([nf-core/nf-co.re#694](https://github.com/nf-core/nf-co.re/pull/694)):
- Monday: Running dev version with MySQL and demo site with queries.
- Tuesday: Got Slack queries running. 
# PEP Pipeline Schema
- Had a nice discussion about how we could potentially go forward with this
- Will use nf-core/rnaseq as a test bed
- Will try to replace validation Python script with some PEP tools
- Idea is to use PEP to validate sample sheet inputs and print out in a standardised "flat" format. This way we can make use of the more advanced PEP features with very little change to the nf-core pipelines.
- _[Nathan]_ Day 2: Wrote a new package to convert a PEP into customizable outputs for downstream consumption, like a project-level YAML format: https://github.com/pepkit/pepconvert
# Refgenie integration
- _[Nathan]:_ Developed a new set of API endpoints to serve unarchived assets via `http` or `s3`, getting us a step closer to using refgenie to replace the aws iGenomes files
- _[Phil]:_ Had GitHub issue throughout all 3 days about RefGenie integration but never quite got to starting it :see_no_evil: