Skip to contents

blisa 1.0.4

  • New element n_cells_by_group in the output of hexBinCells() when group is supplied, giving the number of cells of each cell type in each bin.

blisa 1.0.3

CRAN release: 2026-10-30

  • New argument weighted in plotCCIspatial(). The ligand contribution of each sender cell type at a hotspot is now the spatially weighted sum over its neighbours.
  • New argument weighted in runCCI(). The global ligand count of each sender group is now the sum over hotspots of the local weighted ligand counts used by plotCCIspatial().

blisa 1.0.2

  • Added functionality for visium data
  • New function blisaPathway() for summarizing blisa results at the pathway level
  • Added a new element level to the blisa object class, specifying whether the results are lr level or pathway level
  • plotLRrank() and plotHotspots() not works for pathway level results

blisa 1.0.1

  • New argument platform in blisa to handle xenium and visium
  • New argument as_points in plotHotspots() to plot dots instead of hexagons
  • New argument size in plotHotspots() to change the size of the dots
  • New argument background in plotHotspots() to add background image (e.g., H&E)

blisa 1.0.0

CRAN release: 2026-07-10

  • blisa() now uses the fastLISA C/OpenMP backend by default (fast = TRUE) for the bivariate local Moran’s I computation. fastLISA moved from Suggests to Imports. Results are statistically equivalent to the previous spdep::localmoran_bv backend but not bit-identical (different RNG).
  • New min_total_counts argument to blisa() and hexBinCells() drops bins whose total counts (summed over all genes) fall below the threshold during binning. Default 10.
  • New verbose argument (default FALSE) to blisa() and hexBinCells() gates progress messages; blisa() also shows a progress bar over the ligand-receptor pairs in interactive sessions.
  • New lr_pairs argument to plotCCI() selects specific ligand-receptor pairs to display, overriding top_lr.
  • New main argument to plotCCI(), plotCCILR(), and plotCCIsummary() adds a title above the heatmap.

blisa 0.2.3

  • LR_results row names are now informative IDs built from the ligand/receptor symbols, with subunits joined by | and the two sides by _ (e.g. TGFB1_TGFBR2|TGFBR1). The original CellChatDB interaction_name is retained in a new interaction_name column.
  • plotCCIspatial() now handles multi-subunit ligand/receptor complexes (previously errored with subscript out of bounds).
  • plotHotspots(), plotCCIspatial(), and plotCCILR() now match the ligand/receptor arguments by subunit set, independent of order, separator (,, _, |), and whitespace.

blisa 0.2.2

  • Added fast = TRUE option to blisa() which uses fastLISA::local_moran_bv (C/OpenMP backend) instead of spdep::localmoran_bv for the bivariate local Moran’s I computation. Default behaviour unchanged.
  • New cpu_threads argument controlling parallelism when fast = TRUE.

blisa 0.2.1

  • Update vignettes.

blisa 0.2.0

CRAN release: 2026-06-02

  • Initial CRAN release.