LoginSign Up
← Back to all sparks
M

miscmetabar

ANALYTICS
Velocity0.0

R package miscmetabar by adrientaudiere — release notes from GitHub.

github.com/adrientaudiere/miscmetabar ↗

A phyloseq toolbox keeps absorbing metabarcoding methods, and has started splitting into a family.

metabarcodingphyloseqbioinformaticstaxonomydiversity-analysisr-package
◆Current state
MiscMetabar extends the phyloseq ecosystem for metabarcoding analysis — taxonomic assignment, diversity curves, filtering and plotting for amplicon sequencing data. Recent substantive releases added IDTAXA-based assignment via DECIPHER, UMAP dimensionality reduction, Hill diversity accumulation curves and kmer complexity plots. The most recent release is a CRAN submission bundling several development merges, including a fix for ggplot2 v4.0.0 breaking the upset plot.
◆Where it's heading
The package grows by adopting established methods from elsewhere in the bioinformatics stack rather than inventing its own, wrapping them for phyloseq objects. The 0.11.0 notes floated a pq-verse — splitting functionality into companion packages such as comparpq for comparing phyloseq objects with different taxonomies, explicitly to make maintenance easier. Several releases in this window carry no notes beyond a merge commit, so the feed understates the work behind them.
◆Prediction
The pq-verse split was described as an intention rather than delivered here, so companion packages appearing alongside a slimmer core is the most likely next structural move. Continued ggplot2 v4 compatibility work is the near-term certainty.

◆Recent moves

  1. 10mo ago

    CRAN submission bundling three dev cycles and a ggplot2 v4 fix

    A release note that is a pull request list rather than a changelog, covering everything merged since 0.14. The substantive items are the ggplot2 v4.0.0 fix for upset_plot and a pass of parameter-name harmonisation and memory optimisation.

    View source ↗
  2. 1y ago

    v 0.14.0

    The release body is a merge commit from the development branch with no notes, so what changed at 0.14 cannot be read from the feed. The later CRAN submission suggests real work landed here, but this entry does not describe it.

    View source ↗
  3. 1y ago

    v 0.13.0

    Another merge-commit release with no accompanying notes. Nothing about the contents of 0.13 is visible in the feed.

    View source ↗
  4. 1y ago

    IDTAXA taxonomic assignment added alongside existing methods

    Adding assign_idtaxa() and learn_idtaxa() brings DECIPHER's algorithm in as another option for a task the package already handled, wired through add_new_taxonomy_pq() as a method choice. The sintax changes address a practical annoyance — reference databases with inconsistent rank counts.

    View source ↗
  5. 1y ago

    UMAP, Hill curves and kmer complexity plots added

    The broadest feature release in this window, extending the analysis surface with UMAP ordination, Hill diversity accumulation curves and sequence complexity plots. It also floats the pq-verse idea — companion packages such as comparpq taking on specialised work so the core stays maintainable — which is stated as a direction rather than shipped here.

    View source ↗
  6. 1y ago

    v 0.10.1

    A merge-commit release with no notes, giving no visibility into what changed at 0.10.1.

    View source ↗