Xlavir

Fixes and changes from PR [15](https://github.com/peterk87/xlavir/issues/15)

Fixes

* low coverage coordinate output off by one (`xlavir.tools.mosdepth.get_interval_coords_bed`)
* error on no Pangolin reports found (e.g. non-SARS-CoV-2 report) (`xlavir.tools.pangolin.get_info`)
* user QC thresholds not being used (`xlavir.xlavir.run`)
* not showing all QC fail comments (`xlavir.qc.create_qc_stats_dataframe`)
* consensus sequences being too long for Excel cell character limit (32,767 characters); longer sequences are chunked
into 80 character segments with one segment per line in consensus sheet (`xlavir.tools.consensus.read_fasta`)

Changes

* Ignore and skip unsupported VCFs instead of throwing NotImplementedError (`xlavir.tools.variants.get_info`)
* In consensus sheet, only add QC comments on FASTA header rows if necessary (`xlavir.io.xl.add_comments`)

* Fixed issue ([12](https://github.com/peterk87/xlavir/issues/12)) where iVar ref allele depth corresponds to depth of
base before deletion. For indels, ref allele depth is taken from the total depth minus the alt allele depth.
* Fixed issue ([14](https://github.com/peterk87/xlavir/issues/14)) where the total number of reads
from `samtools flagstat` may not be the true number of reads. The unmapped reads may be excluded from the BAM file so
the `samtools flagstat` total number of reads may be equal to the number of mapped reads. There is now a search for
fastp JSON files to get the true total number of reads.

* Added support for Nanopolish VCF parsing as generated by the ARTIC pipeline
* Added deduplication of VCF and SnpSift entries since the ARTIC pipeline may produce VCF files with duplicate variant
calls due to overlap between amplicons.
* Added VCF and SnpSift test data for CLI test to generate Excel report.

* Fix an issue where single base positions are being reported as 0-based when all other ranges are 1-based for reporting
of low/no coverage regions from Mosdepth per-base BED files ([10](https://github.com/peterk87/xlavir/pull/10)).

* Add support for nf-core/viralrecon version 2.0 (requires Mosdepth `bed.gz` files be output; needs
custom `modules.config` like [this one](https://gist.github.com/peterk87/495621349c1161d12047c1c8f97935af))
* [Nextclade CLI](https://github.com/nextstrain/nextclade/blob/master/packages/cli/README.md) per sample results parsed
into sheet showing useful info like Nextstrain clade, of mutations, of PCR primer changes
* Added check that input directory exists and is a directory
* Added sheet with xlavir info
* Added Gene, Variant Effect, Variant Impact, Amino Acid Change to Variant Summary table

* Add reference sequence length to QC stats table. Get ref seq length from max mosdepth per base BED coverage value.
* Add more conditional formatting
* Fix `execution_report.html` finding
* Fix version printing; add to help
* Add epilog with usage info