the StaPH-B call on 28 February 2022 (video) (slides)
Considerations:
(easy) freyja update generates a barcode file from the latest (daily) UShER tree. This necessitates a separate "setup" process very early in the workflow.
From what point in the pipeline do we want to feed data to freyja variants? My guess would be "after SAMTOOLS_SORT_INDEX", because that produces the BAM files it needs to produce the VCF output file and the sequencing depth file.
Related: do we want to replace the outputs from the IVAR_VARIANTS step with the future freyja variants step?
After freyja demix, we might want to run freyja aggregate followed by freyja plot. However, as far as I can tell, Nextflow currently has no way of passing a single directory to a process containing the aggregate output of multiple processes (but I may be wrong). The easy way is to submit a fix to upstream freyja which lets freyja aggregate also work on specified individual files.
Resources:
Considerations:
freyja updategenerates a barcode file from the latest (daily) UShER tree. This necessitates a separate "setup" process very early in the workflow.freyja variants? My guess would be "afterSAMTOOLS_SORT_INDEX", because that produces the BAM files it needs to produce the VCF output file and the sequencing depth file.IVAR_VARIANTSstep with the futurefreyja variantsstep?freyja demix, we might want to runfreyja aggregatefollowed byfreyja plot. However, as far as I can tell, Nextflow currently has no way of passing a single directory to a process containing the aggregate output of multiple processes (but I may be wrong). The easy way is to submit a fix to upstream freyja which letsfreyja aggregatealso work on specified individual files.