All notable changes to this project will be documented in this file. This project adheres to Semantic Versioning.
- Abundance list reported by
kevlar filter
now correctly show re-computed proband k-mer abundances, not pre-filtering abundances as before (see #111). - The
kevlar localize
andkevlar call
procedures now handle multiple assembled contigs and multiple reference matches (see #124 and #126).
- New abundance screen now a part of
kevlar novel
. If any k-mer in a read is below some abundance threshold, the entire read is discarded (see #106). - Better error reporting and handling of various issues with assembly, localization, and alignment (see #113, #114).
- Preliminary support for VCF output.
- The
kevlar filter
procedure now handles both contamination and reference matches under a single mask interface (see #103). - Explicitly dropped support for Python 2.7. Now supports only Python >=3.5 (see #125).
- The
kevlar collect
command and associated tests. Its functionality has now been fully distributed to other subcommands.- Read filtering to
kevlar filter
- Junction count contig assembly to
kevlar filter
as an optional mode
- Read filtering to
- New subcommands
partition
: group reads by shared interesting k-merslocalize
: determine an assembled contig's location in the reference genomecall
: align assembled contigs to reference and call variant
- Documentation suite in
docs/
, hosted at https://kevlar.readthedocs.io - New third-party dependency
ksw2
for computing alignments. Wrapped with Cython, which is a new development-time dependency (but not install or run time). - The
pandas
package is now a dependency, andpysam
andnetworkx
are now hard dependencies (rather than conditional).
- Bug with assembly when the order of a read pair was swapped and they had the opposite orientation (see #85).
- Command-line interface with 8 subcommands
dump
: discard reads that match reference completelycount
: compute k-mer abundances for all samplesnovel
: identify "interesting" (potentially novel) k-mersfilter
: re-compute k-mer abundances, discard false positives and contaminationassemble
: assemble reads for a single variantcollect
: collect and filter (legacy)mutate
: simulate variants on a genomereaugment
: re-attach interesting k-mer annotations to reads
- Extensive test suite
- Continuous integration configuration