v1.0.0-beta

First major release of the pipeline.

1. Added nf-test testing
2. Added support for real-time (live) analysis of clusters per sample during sequencing
3. Added GPU acceleration for medaka polishing
4. Split the workflow into several subworkflows
5. Implemented batching of input files for higher parallelization
6. Implemented a process to glue molecule files before submission to medaka polishing

• to decrease number of files to process, which can be up to several 100k for large sequencing experiments
• to make better use of the GPU acceleration

7. Added support for multi-line bed files to process several targets in one run

• Output files are therefore split per target

8. Added simple plots and stats to keep track of the number of clusters per sample

CAVEAT: GPU acceleration supporting docker container is only available up to v2.0.0, in which medaka smolecule is broken. v2.0.0 uses the new pytorch models, and therefore, CPU medaka and GPU medaka results might differ slightly.
The Docker container needs to be built manually since there are problems with quay.io.

0.2.1

• Version of vcflib now specified (v1.0.0)
• Version 1.0.9 of vcflib caused issues running with freebayes
• Next release will include automated testing

0.2.0

• Release of the updated clustering strategy
• Support of singularity improved
• Dynamic resource allocation upon retries for cluster polishing
• Release is up to date with the paper pipeline version

0.1.0

• Init release of the UMI pipeline
• Pipeline is production ready and comes with a server-hosted docker container