broadinstitute/viral-ngs: v1.12.0

New:

• New BWA-based LCA metagenomics function: metagenomics. align_rna()
  • Align to metagenomics bwa index, mark duplicates, and generate LCA report
• In Broad Snakemake pipeline, problematic nodes can now be blacklisted
• easy-deploy-viral-ngs.sh now supports Python 2-based conda environments (sans Snakemake), and installation of viral-ngs from local conda builds
• GATK_PATH (the path to the directory containing the GenomeAnalysisTK.jar file) can be specified via --GATK_PATH to functions requiring GATK
• NOVOALIGN_LICENSE_PATH (the path to a novoalign.lic file) can be specified via --NOVOALIGN_LICENSE_PATH to functions that make use of Novoalign

Fixed:

• reports.rules Snakemake file fixes usage of samples-runs.txt
• easy-deploy-viral-ngs.sh has bugfix to remove debug lines from before the file had automated testing.
• Fix bug in calculating coverage when mapped bam has no reads: coverage quantiles are only reported when there are mapped reads
• Fix bug in tools.bwa.align_mem_one_rg() (used by reports.align_and_plot-coverage()) where input bam was removed if the file had only one RG
• assembly.refine_assembly more robust to empty input

Changed/Updated:

• easy-deploy-viral-ngs.sh script now displays a shorter command prompt (modified $PS1)
• easy-deploy-viral-ngs.sh script now tests for 768MB RAM to account for lower ish limit on Broad systems
• In Snakemake pipeline, Kraken results are also now reported for cleaned reads
• Testing and deployment-related files can now be found in broadinstitute/viral-ngs-deploy. The files will be leaving this repository in the near-ish future
• pysam 0.8.3 -> 0.9.1
• picard 1.141 -> 2.5.0
• diamond 0.7.10 -> 0.8.22
• snakemake 3.7.1 -> 3.8.2
• pip packages moved to requirements-conda.txt. This mostly impacts testing, since the dependencies are listed elsewhere for the bioconda package of viral-ngs