Version 0.3.0

Major updates since v0.2.0

• Count annotations: represent k-mer counts, encode abundances or gene expression

  • query count annotation in #309
  • delta-coding of k-mer count annotations in #315
  • query k-mer count quantiles in #331

• Coordinate annotations: represent k-mer coordinates (all occurrences in the source) in multi-integer annotations (=> lossless indexing of sequences, encoding of sequence traces)

  • index and query k-mer coordinates in #326, #337, #338, #341, #345
  • coordinate annotations with Multi-BRWT base in #353

Other new features:

• window smoothing of k-mer counts within unitigs in #316
• print histogram of k-mer weights (counts) in #351
• new query parameter: filter results by the number of k-mers matched to the joint graph in #339
• skeleton for taxonomic classification in #343 and #346

Enhancements:

• Alignment refactor and rewrite in #333
• Chain local alignments to improve accuracy of long read alignment in #335
• optimized storage of labels in batch annotator (=> faster queries, reduced RAM) in #350
• improvements in column aggregation in #314
• support row-diff transform in memory without disk swap in #342
• parallel transforms for changing graph state (e.g., small->stat) in #312
• serialize node weights immediately when building graph (=> reduced RAM) in #317
• improved parallelization of annotating sequence batches in #322
• bit_vector_sd::conditional_rank1 for faster DBGBitmap::map_to_nodes in #323
• various other tweaks and cleanups
• sphinx documentation

Deployment:

• docker containers are now published on ghcr.io (see https://github.com/ratschlab/metagraph/pkgs/container/metagraph)

Full Changelog: v0.2.0...v0.3.0