Changes in version 2.2.1.9002                      

  - Improved scalability of downstream analysis and visualization
      - Reduced the need of pre-calculated normalized data for
        performing Wilcoxon's test and producing feature expression
        plots
      - Normalized data will only be calculated on the fly from raw data
        and pre-stored size factor (obj$nUMI).
  - Added theme_axis_shortArrow() for tidy dimensional reduction plot
    axis theme.
  - Migrating to patchwork for multi-plot layouting. Mainly for the ease
    of alignment, legend collection, and subplot extraction.

                 Changes in version 2.2.1 (2025-08-26)                  

  - Added naive GSEA analysis on factor gene loading (W) to test if any
    known gene sets (e.g. cell cycle) is enriched in any factor.
    Implemented in factorGSEA().
  - Added dense data loading support for H5AD files
  - Optimized obs metadata parsing for H5AD files
  - Fixed ggplot2 color picking when coloring by logical value
  - Fixed H5AD file layer detecting bug
  - Fixed some other minor bugs

                 Changes in version 2.2.0 (2025-05-01)                  

  - Implemented highly efficient on-disk iNMF that scales to a million
    cells using slightly more time than in-memory version, requiring
    only laptop-level memory.
  - Added 10X H5 data and H5AD loading function that loads the data into
    regular dgCMatrix in memory or the DelayedArray representation
    backed on disk, the latter is used for on-disk iNMF implementation.
  - Added selectBatchHVG() which implements another HVG selection
    strategy, credit to SCIB
  - Adding suggestK() back with new methodology
  - Clarified optimal runGOEnrich() workflow and added fold enrichment
    metric in the returned result
  - Fixed important bug in online iNMF scenario 2
  - Fixed multiple problems related to ATAC analysis
      - Fixed Wilcoxon rank-sum test bug when using ATAC peak counts
      - Fixed gene coordinate parsing bug from BED file
      - Optimized peak parsing speed

                 Changes in version 2.1.0 (2024-10-29)                  

  - Added centroidAlign() for new cell factor loading alignment method
  - Added plotProportionBox() for visualizing compositional analysis
  - Added plotClusterGeneViolin() for visualizing gene expression in
    clusters
  - Added plotBarcodeRank() for basic QC visualization
  - Added plotPairwiseDEGHeatmap() for visualizing pairwise DEG results
  - Added plotGODot() for visualizing GO enrichment results
  - Added calcNMI() for evaluating clustering results against ground
    truth
  - Added ligerToH5AD() allowing reticulate/Python free export of liger
    object to H5AD format. This is presented in extension source code
    (i.e. not loaded with library(rliger)).
  - Added organism support in runGeneralQC() and refined hemoglobin gene
    matching regex pattern.
  - Optimized DE test memory usage scalability for both pseudo-bulk
    method and wilcoxon test
  - Optimized plotProportionPie() by adding argument circleColors
  - Optimized plotVolcano() text annotation positioning and gene
    highlighting logic.
  - Optimized visualization function additional argument documentation
  - Changed runMarkerDEG() and runPairwiseDEG() default method from
    "wilcoxon" to "pseudoBulk"
  - Fixed runMarkerDEG(method = "pseudobulk") bug in assigning
    pseudo-replicates, and optimized error/warning signaling.
  - Fixed bug in calcAlignment(), subsetMemLigerDataset(), cellMeta()
  - Fixed bug in old version updating functions

                 Changes in version 2.0.1 (2024-04-04)                  

  - Fixed wrong UINMF aborting criteria
  - Fixed example/test skipping criteria for non-existing dependencies
  - Fixed file access issue when checking on CRAN
  - Updated installed data file system.file("extdata/ctrl.h5",
    "extdata/stim.h5") to be of standard 10X H5 format
  - Updated quantileNorm() automatic reference selection according to
    #297
  - Other minor fixes (including #308)

                 Changes in version 2.0.0 (2024-03-20)                  

  - Added ligerDataset class for per-dataset information storage, with
    inheritance for specific modalities
  - Added a number of plotting functions with clear function names and
    useful functionality
  - Added Leiden clustering method, now as default rather than Louvain
  - Added pseudo-bulk DEG method
  - Added DEG analysis with one-vs-rest marker detection in
    runMarkerDEG() and pairwise comparison in runPairwiseDEG()
  - Added gene name pattern for expression percentage QC metric
  - Added native Seurat object support for the core integration workflow
  - Added a documentation website built with pkgdown
  - Added new iNMF variant method, consensus iNMF (c-iNMF), in
    runCINMF(). Not stable.
  - Added GO enrichment dowsntream analysis in runGOEnrich()
  - Changed liger object class structure
  - Moved iNMF (previously optimizeALS()), UINMF (previously
    optimizeALS(unshared = TRUE)) and online iNMF (previously
    online_iNMF()) implementation to new package RcppPlanc with vastly
    improved performance. Now wrapped in runINMF(), runUINMF() and
    runOnlineINMF() respectively, and all can be invoked with
    runIntegration().
  - Updated H5AD support to match up with Python anndata package 0.8.0
    specs
  - Renamed many function/argument names to follow camelCase style,
    original names are still available while deprecation warnings are
    issued

                 Changes in version 1.0.1 (2023-11-08)                  

  - Allow setting mito pattern in getMitoProportion() #271
  - Fix efficiency issue when taking the log of norm.data (e.g.
    runWilcoxon)
  - Add runable examples to all exported functions when possible
  - Fix typo in online_iNMF matrix initialization
  - Adapt to Seurat5
  - Other minor fixes