NEWS.md

Release Notes

• 1.7.0

  • Spike-in methodology has been overhauled. Now, reads will be aligned to a combined genome of the reference + all spike-in genomes. Filtering & Duplicate marking are done on the combined alignment file which is subsequently split into separate bam files for each genome after filtering. Based on work by Jeanne-Marie McPherson demonstrating this strategy provides a more accurate estimate of experimental:spike-in reads.
    • The combined genome alignment strategy has a small side effect of adding extra chromosomes to each bam header corresponding to the spike-in genome chromosomes. These shouldn't cause downstream issues, but please file an issue report if this causes unintended breakage.
    • Default align step runtime increased to 1:30:00 from 1:00:00 to account for larger genomes
  • Users can now set spikeGenome to an array of genome entries to perform spike-in normalization to multiple genomes
    • To support these changes, spike-in normalized bigwig files are now denoted: "{species}-spikeNorm.bw" instead of merely "spikeNorm.bw"
    • The output sampleSheet columns referencing spikeNorm files now also contain the spike-in genome name i.e. "bigwig_allFrags_{species}_spikeNorm" vs old "bigwig_allFrags_spikeNorm"
  • No longer uses temp() anywhere, intermediate files will be kept at all steps. These steps are candidates for merging in the future to eliminate the need to keep intermediates.
  • Computes genome alignment stats in: Logs/{sample}*_genomeStats.tsv
    • added local software dependency: mlr, set in mlrPath in config.json
  • Pipeline now uses envmodules (no more manual calling module purge & load), and --use-envmodules is now set in slurmSubmission.sh. This enables easier adaptation of other module management systems like conda, docker & singularity.

• 1.6.4

  • Added computation of alignment stats
  • No longer uses temp() for some intermediate files.