Skip to contents

Changelog

Source: NEWS.md

comparpq 0.2.1 (Development version)

  • refseq_comp_lpq() new function to compare @refseq sequences across all phyloseq objects in a list_phyloseq using k-mer Jaccard similarity and union-find connected components. Returns per-threshold Venn diagrams and shared-cluster counts. No igraph dependency.

  • find_primers_pq() new function to detect taxa whose reference sequences match primer sequences (IUPAC-aware, forward and reverse complement). Returns a data frame suitable for use with tidypq::filter_taxa_pq() to prune contaminated taxa.

  • community_sharing_barplot_pq() new function to display pairwise community-sharing metrics as grouped bar charts, faceted by metric or by pair. Companion to community_sharing_pq().

  • community_sharing_pq() new function to visualize community sharing between 2–4 modalities of a sample variable as a network figure: each node is a pie chart of taxonomic composition, and curved links encode multiple pairwise similarity metrics (Bray-Curtis, Jaccard, shared species, shared genera proportion). Supports label-permutation significance testing (n_perm). Requires packages ggforce and RColorBrewer.

  • default_sharing_metrics() new function returning the 4 built-in metric definitions used by community_sharing_pq() and community_sharing_barplot_pq().

  • make_sharing_metric() new function to create custom metric definitions for community_sharing_pq() and community_sharing_barplot_pq().

  • div_pq() and hill_samples_pq() now use divent (via MiscMetabar::divent_hill_matrix_pq()) instead of vegan::renyi() for Hill number computation, and divent::ent_shannon() / divent::ent_simpson() instead of vegan::diversity() for Shannon and Simpson indices. The default estimator is now "UnveilJ" (bias-corrected); pass estimator = "naive" to restore old numeric behavior.

  • div_pq(): the scales parameter is deprecated in favour of q. The hill parameter is deprecated; only Hill numbers are now supported.

  • add_external_seq_pq() now checks for a refseq slot upfront and emits a clear error when absent, instead of crashing with a cryptic message. It also strips the phy_tree slot before calling merge_phyloseq() to avoid tip-count mismatches on objects that carry a tree.

  • add_shuffle_seq_pq() now checks for a refseq slot upfront and emits a clear error when absent. It also strips the phy_tree slot before calling merge_phyloseq() to avoid tip-count mismatches.

  • compare_refseq() correctly handles list_phyloseq S7 objects by accessing @phyloseq_list directly instead of calling length() on the S7 object itself.

  • estim_cor_pq() / estim_cor_lpq() bootstrap now passes use = "complete.obs" to stats::cor() and na.rm = TRUE to stats::quantile(), preventing NaN-induced crashes on degenerate resamples.

  • estim_diff_pq() now validates that each group has at least 3 samples before delegating to dabestr, providing an informative error message instead of a cryptic dabestr crash.

  • rainplot_taxo_na() now checks that requested rank columns exist in the psmelt() output before calling across(), providing a clear error when all-NA rank columns are dropped.

  • tc_heatmap() new function to visualize the correspondence between two taxonomic ranks as a heatmap, where each cell shows the number of taxa assigned to a given pair of rank values.

  • taxtab_replace_pattern_by_NA() fixes an inner-loop variable bug where patterns were applied to all taxonomic_ranks columns simultaneously instead of one at a time.

  • tc_points_matrix() now checks that requested rank columns exist in the psmelt() output before grouping, providing a clear error when all-NA rank columns are dropped.

  • Add param compute_dist to list_phyloseq()

  • length(), names(), [(), and [[() now work correctly on list_phyloseq objects. S7 stores the class attribute as "comparpq::list_phyloseq" (package-qualified), which prevented S3 dispatch from finding length.list_phyloseq. The constructor now prepends the bare "list_phyloseq" name to the class vector, enabling S3 dispatch. The four accessor methods are now documented and exported.

comparpq 0.1.2

  • Add params significance, test and p_alpha to div_pq() to report tuckey hsd paired-test using letters.
  • gg_hill_lpq() new function to visualize Hill diversity correlations across pairs of phyloseq objects in a list_phyloseq. Produces a faceted scatter plot (pairs × Hill orders) with optional 1:1 line, regression line, and per-panel correlation annotation, enabling visual assessment of REPRODUCIBILITY, ROBUSTNESS, and REPLICABILITY.
  • gg_bubbles_pq() now uses merge_lpq() when a list_phyloseq is passed, merging into a single phyloseq and faceting by source_name instead of building separate patchwork panels. diff_contour now works with any facet_by variable (not just list_phyloseq), highlighting taxa unique to each facet level with a distinct contour color from diff_contour_colors. No longer limited to 2 or 3 objects. New match_by parameter controls how taxa are matched when merging list_phyloseq objects.
  • gg_bubbles_pq() new ggplot2-based circle-packed bubble plot of taxa abundances. Unlike bubbles_pq(), it does not require d3js/Observable and supports faceting by a @sam_data variable to display one bubble chart per level. Uses packcircles for layout computation.
  • compare_refseq() new function to compare reference sequences (refseq slot) between two phyloseq objects, identifying shared and unique ASVs/OTUs by name and by DNA sequence content, including detection of same-name-different-sequence and same-sequence-different-name mismatches. Computes mean nearest-neighbor k-mer distance for unique sequences.
  • Add function apply_to_lpq() to apply a function to each phyloseq object in a list_phyloseq
  • estim_cor_lpq() new function to compute bootstrap correlation/regression across a list_phyloseq
  • estim_cor_pq() new function to compute bootstrap correlation and regression CIs for diversity vs numeric variables
  • estim_diff_lpq() new function to run estimation statistics (effect sizes + CIs) across a list_phyloseq
  • estim_diff_pq() new function for estimation statistics (Gardner-Altman/Cumming plots) comparing diversity across groups via dabestr
  • merge_lpq() new function to merge a list_phyloseq into a single phyloseq object where each original phyloseq becomes one sample. Taxa can be matched by reference sequences (match_by = "refseq", default) or by taxa names (match_by = "names").
  • simple_venn_pq() new function to draw Venn diagrams of shared taxa across 2-4 sample groups using pure ggplot2 (no external Venn package needed), with support for multiple taxonomic ranks and compact, clearly labeled circles/ellipses.
  • simple_venn_pq() now accepts a list_phyloseq object as input, automatically merging it via merge_lpq() before drawing the Venn diagram.