Kevlar version 0.7

Added

• A new Snakemake workflow for preprocessing BAM inputs for analysis with kevlar (see #305, #355).
• A new Snakemake workflow for kevlar's standard processing procedure (see #306, #355).
• New unband module to merge augmented Fastq files produced with a k-mer banding strategy (see #316).
• New varfilter module to filter out preliminary variant calls overlapping with problematic/unwanted loci or features (see #318, #342, #354).
• New dependency: intervaltree package (see #318).
• A new sandbox directory with convenience scripts for development and analysis (see #335).
• A new --min-like-score filter for the simlike module (see #343).
• A new --drop-outliers filter for the simlike module (see #350).

Changed

• Added a new flag to print to the terminal (stderr) and a logfile simultaneously (see #308).
• The functionality of the previous filter module is now split between the new unband module and a reimplementation of the filter module (see #316).
• Added a "fast mode" to the simlike module, prematurely halting computations for calls already marked for filtering (see #328).
• Added a filter for problematic short indels adjacent to homopolymers (see #336, #338, #339).
• Implemented new filters in the simlike module based on thresholds and k-mer abundances: the ControlAbundance filter for predictions with too many high-abundance parent/control k-mers spanning the variant, and the CaseAbundance filter for predictions with too many consecutive proband/child k-mers spanning the variant (see #327, #339).

Fixed

• Corrected a bug that reported the reference target sequence instead of the assembled contig sequence in the CONTIG attribute of indel calls in the VCF (see #304).
• Corrected a bug that called adjacent substitutions as independent SNVs rather than an aggregate MNV (see #332).

Removed

• The effcount, dump, and simplex modules have been dropped (see #308, #316).
• Internal handling of interesting read mate sequences has been dropped (see #353).