v1.3.4

• Add support for samplesheet input mode using --samplesheet_input flag (Issue #53)

v1.3.3

Fixes

• Fixed issue where pipeline would crash when /tmp directory filled during downsampledBamToFastq process (Issue #46)

v1.3.2

• Ensure that an alignment & consensus are generated for all samples, even in the case where all reads are filtered out by trimming & dehosting (Issue #44)

v1.3.1

Minor changes:

• Change extension of freebayes-generated consensus from .fasta to .fa for consistency with our other consensus sequences (Issue #36)
• Remove .primertrimmed from the filenames of output files downstream of the freebayes-generated consensus (Issue #37)
• Remove some config files that are not needed for our execution environment (Issue #38)

v1.3.0

• Use freebayes for consensus generation and variant calling (Issue #32). Output directory is ncovIllumina_sequenceAnalysis_callConsensusFreebayes.
  • also run bcftools csq to annotate .vcf files produced by freebayes. Output directory is ncovIllumina_sequenceAnalysis_annotateVariantsVCF. Both .vcf and .tsv outputs are created.
  • Note: ivar variants and ivar consensus are still run, but the freebayes consensus is carried forward for all downstream analysis.
• Incorporate a GFF annotation file for use with ivar variants and bcftools csq (Issue #31). GFF file is optional. If the GFF file is not provided, bcftools csq will not be run.
• Add two additional fields to ivar variants output: CODON_POS and MUT_NAME. Codons are determined using the GFF file and the ivar_variants_add_codon_position.py script. When no GFF file is provided, these fields are set to NA. Output directory is ncovIllumina_sequenceAnalysis_addCodonPositionToVariants.

v1.2.0

• Add a 'downsampling' step after reads are aligned & primertrimmed. Included to guard against excess depth of coverage when using NextSeq 2000 platform (Issue #29)
• fastq.gz files set aside for NML upload are based on reads from downsampled alignments.
• A downsampling summary is written alongside the QC summary.
• Modified default value for mpileupDepth parameter to be consistent with oicr-gsi/ncov2019-artic-nf.

v1.1.1

• Bugfix: Handle case in filter_non_human_reads.py where total_reads = 0, which would otherwise lead to DivideByZeroError and full pipeline failure. (Issue #26, Pull-Request #28)

v1.1

• Perform dehosting upstream of analysis
• Update to ivar-1.3 and incorporate amplicon filtering
• Publish dehosted .fastq.gz and consensus files to an nml_upload directory
• Routinely align each consensus against the reference, and trim 5' and 3' UTR regions
• Index all published .bam files

v1.0

Remove mpileup depth limit by default (#63)