CHANGELOG

v1.4.5:
    * Fieldbioinformatics now supports rapid barcoded (fragmented) primer trimming and normalisation
    * Nanopolish has been removed completely due to several compatibility issues
    * Medaka has also been removed completely due to kicking out long indels in a way that cannot be changed.
    * Clair3 is now the default variant caller, by default only the r9.4.1 models are available but a artic_get_models command has been added which will fetch the ONT created r10.4.1 models listed in the rerio repository.
    * The pipeline will also attempt to pick an appropriate model based on the basecall_model_version_id field that is added to read headers by default by ONT sequencers.
    * Removed longshot entirely, it also kicks out long variants and is now unnecessary due to clair3 being a much better variant caller.
    * Primer scheme fetcher has been updated to pull from the quick-lab primal hub schemes repository. For schemes not available in this repository you may provide them directly with the arguments --bed and --ref.
    * Automated docker builds pushing to quay.io for use in nextflow pipelines etc.
    * Remove some old functionality which is no longer relevant (basecalling, gather, etc)
    * Re-implement CI as a gh action.
    * Fix the overlapping variants issue by normalising variants against the pre-consensus using bcftools norm.

v1.1.0-rc1:
    * Support for read groups:
        * Support ‘pool’ read groups taken from BED file, e.g.:
            * nCoV2019_1
            * nCoV2019_2
        * These can be helpfully viewed in the Tablet viewer.
        * Do variant calling separately on each group, permits detection of incongruent mutations e.g. from contamination
    * Update to nanopolish 0.12.5 fo more informative VCF output
    * Change nanopore filtering to support new VCF output (e.g. strand-bias)
    * Support indels with bcftools consensus
    * Longshot added to Medaka to permit filter VCF on depth
    * New workflow model - all commands run "one per sample", script nanopolish index
    * Add 'artic gupplyplex' for support for demultiplexed Guppy output to replace 'artic gather'
    * Fix for align_trim to remove erroneous CIGAR strains that can cause medaka to fail
    * Support for amplicons v3 format file (Will Rowe)
    * Test suite (Will Rowe)
    * Documentation (Will Rowe)
    * Remove old deprecated code (Will Rowe)

v0.13:
    * ARTIC fork

v0.12: 
    * Update to nanopolish v0.8.4 for Albacore 2.0+ support:
        * run nanopolish -d /path/to/fast5 input.fasta before zibra.py minion

v0.11-dev:
    * Add stringent pipeline

v0.11 (25th July 2017):
    * Revert back to Ryan Wick's porechop repo as --untrimmed now implemented
    * New script: align_trim_fasta.py and reconstitute.py as basis for more principled demultiplexing (align first, then trim to known amplicon end points +/- 40 bases) to help prevent chimeric reads resulting in incorrect barcode identification

v0.10 (15th July 2017):
    * Better handling of confident deletions (previously ignored, now N-masked)
    * Update to nanopolish HEAD
    * Fixes to support latest nanopolish VCF format changes
    * Ability to choose higher values of max-haplotypes to support more divergent references

v0.9:
    * New command line interface zibra.py to ease extraction and demultiplexing with Albacore and multiple groups

v0.8:
    * Track latest Porechop with better adaptor trimming

v0.7:
    * Add Illumina wrapper scripts

v0.6:
    * Update poretools

v0.5:
    * Update nanopolish to latest HEAD - to fix Albacore 1.1 support (7de633d01cc35a58e5537af5dd1024ae0040d15c)
    * Add Lassa virus scheme

v0.4:
    * Update nanopolish to latest HEAD - aaaf90beb4efe37243c651e62fcfd11374fe2176 for Albacore 1 support
    * Initial support for Yellow Fever (500 and 1000 bp schemes)
        * Move to poretools for extraction with named basecaller

v0.3:
    * add Porechop

v0.2:
    * update nanopolish to 0.6 - adds default support for various nanopore models
    * update samtools to 1.4