Plot_Hierarchy_Proportion() now accepts
adjust_by_group = TRUE to scale per‑group proportions by
relative cell counts, highlighting both within‑group composition and
between‑group abundance differences. Heatmap row labels now also include
the total cell count per group.CellTypist: pre‑computed
markers for 399 cell types across 12 human organs, derived from the
CellTypist organ atlas (Xu et al. 2023; Domínguez Conde et
al. 2022). Markers were identified using the Scanpy workflow
(Wilcoxon test, adjusted p‑value < 0.01, log2 fold‑change > 0),
ranked by log fold‑change, and cut to the top 100 genes per cell type.
See ?CellTypist and the package README for usage
details.Plot_Hierarchy_Proportion()."Markers_list_scIBD" dataset has been re‑calculated
using the Scanpy sc.tl.rank_genes_groups function; its
documentation and the README have been updated to reflect the new
workflow.'paletteDiscrete' function
failed to be called due to the absence of the '@export'
attribute.Plot_Hierarchy_Proportion() creates a composite figure
with a hierarchical cell‑type tree (bubbles and trapezoids) and an
optional group proportion heatmap, handling partial sub‑type annotations
and providing ArchR‑based or fallback colour palettes.Plot_Voronoi_diagram() generates a weighted Voronoi
treemap of hierarchical cell‑type composition, with polygon area
proportional to cell count and sub‑type colours matching the same
palette logic as Plot_Hierarchy_Proportion.paletteDiscrete() function and
ArchRPalettes dataset that faithfully reproduce the colour
palettes (stallion, calm, kelly, etc.) from the ArchR package. The
function supports natural sorting of category names, automatic
interpolation when the number of categories exceeds the palette size,
and reversal of colour order.Plot_Hierarchy_Proportion function for hierarchical cell
type visualisation.Plot_Voronoi_diagram with
a detailed parameter guide and example code.Compute_Gene_AUC_ROC() computes single‑gene AUC and
optionally plots the ROC curve, supporting raw and rank‑based scoring,
expression truncation, and cell subsetting for robust marker
evaluation.README.md with Section 5.2 demonstrating the
new function.Celltype_Compare(): a robust function for
cross‑tabulating cell type labels and grouping variables between two
single‑cell objects (e.g., Seurat). It automatically aligns cell
barcodes using multiple normalisation strategies, returns count tables,
column‑wise proportion tables, a dominant mapping, and a
publication‑ready heatmap..onAttach.library() calls from
Celltype_Calculate() to comply with R packaging best
practices. All dependencies are now accessed exclusively through
NAMESPACE imports, preventing namespace pollution.geom_line(size = ...) with
geom_line(linewidth = ...) in ROC curve plotting.Markers_filter_ScType() for filtering ScType
markers by tissue type and cell name.ScType, ScType_raw, and
ScType_table data objects.This release introduces a major new feature for fine-grained cell type identification alongside significant improvements to accuracy, performance, and usability.
weighted (default),
mean, and AUCell.min_score = "auto" threshold
that scales with the number of cell types, preventing excessive
“Unassigned” labels in large marker sets (e.g., 30+ types).min_confidence parameter for ratio-based
filtering, providing more robust cell type discrimination than simple
score differences.weighted scoring method with new marker
specificity and IDF-like weights.AUCell method with adaptive
n_top calculation and a combined binary/rank-weighted
scoring strategy.Raw_score_matrix and a
Parameters list for better reproducibility and
debugging.RANN package for optional 10-100x faster
k-NN computations in spatial smoothing.compute_adaptive_parameters().Celltype_Calculate() to
use individual gene AUCs for more robust predictions.Parameter_Calculate() for better model generalization.Parameter_Calculate() to include threshold
parameter prediction.has_colnames parameter to
Read_excel_markers() to support reading Excel files without
column headers.Markers_list_scIBD
database.Read_seurat_markers() for outputs from the
presto package.Celltype_Verification().Celltype_Verification() function for
generating validation dot plots.colour_low,
colour_high) for all plotting functions.Read_seurat_markers() with better
compatibility for FindMarkers results from the
presto package.Celltype_Calculate() function.Celltype_Calculate() function for
automated cluster scoring.Celltype_Annotation() function for an
end-to-end annotation workflow.Celltype_annotation_Heatmap() function.calculate_probability().Celltype_annotation_Heatmap() to use the new
calculate_probability() function.Celltype_annotation_Bar() to
Celltype_annotation_Box() and improved its visualization
output.