Changes in version 1.0.4
SIGNIFICANT USER-VISIBLE CHANGES
- Return message instead of error when there are no cells of interest
present in the image (identify_neighborhoods()).
NOTES
- Moved package xROI from imports to suggests.
Changes in version 1.0.3
BUG FIXES
- Fixed error when there are only one cell in the clusters.
(identify_neighborhoods()).
- The calculation of cell types of interest to
All_cells_in_the_structure in
calculate_proportions_of_cells_in_structure() was incorrect. Now
fixed.
SIGNIFICANT USER-VISIBLE CHANGES
- Re-organised the vignettes.
Changes in version 1.0.2
BUG FIXES
- Fix bug when Cell.ID column is missing from the spe_object in
identify_neighborhood().
Changes in version 1.0.0
SPIAT released on Bioconductor 3.16.
Changes in version 0.99.14
BUG FIXES
- Fixed bugs in identify_neighborhoods():
1) Assign "Free_cell" to the cells of interest when the number of
clustered cells are smaller than min_cluster_size in each cluster;
2) Fixed spe_object output (Adding "Neighborhood" column had a bug
previously).
Changes in version 0.99.13
BUG FIXES
- Minor bug in average_nearest_neighbor_index() - the variable output
was not defined when there is no reference cell found in the image.
Changes in version 0.99.12
SIGNIFICANT USER-VISIBLE CHANGES
- average_nearest_neighbor_index() now returns the index along with
the pattern type and the p value.
Changes in version 0.99.11
BUG FIXES
- Renamed the file of R_BC().
- Fixed a minor bug in identify_bordering_cells() that causes error.
Changes in version 0.99.10
SIGNIFICANT USER-VISIBLE CHANGES
- mixing_score_summary() ensures returning data in any situation.
There will be difference between returning NA and 0. See updated
documentation.
Changes in version 0.99.9
BUG FIXES
- Fixed format_image_to_spe() "general" format. Assigned rownames
(markers) and colnames (Cell IDs) to the assay of the formatted spe
object.
Changes in version 0.99.8
BUG FIXES
- Fixed format_image_to_spe() "general" format. The NA intensity value
removal was for markers. Instead, it should remove cells that have
NA marker intensities.
- Fixed predict_phenotypes() plot error.
Changes in version 0.99.7
BUG FIXES
- Fixed formatting image error when intensity_matrix is NULL under
"general" option.
Changes in version 0.99.6
SIGNIFICANT USER-VISIBLE CHANGES
- identify_bordering_cells() emit Warning when no bordering cells are
detected.
BUG FIXES
- Fixed the message displayed for NA removal when formatting image
using "general" format.
- Fixed plot functions to display only one plot.
plot_marker_level_heatmap(), and plot_distance_heatmap().
Changes in version 0.99.5
SIGNIFICANT USER-VISIBLE CHANGES
- Generalised functions in tumour structure families to suit other
tissue and cell types, not restricted to tumour and immune cells.
The affected functions include:
1. identify_bordering_cells() - deleted "tumour" from the plot
title;
2. plot_cell_categories() - cell categories changed to not contain
"immune" when feature_colname == "Structure";
3. calculate_distance_to_tumour_margin() renamed to
calculate_distance_to_margin();
4. R_BT() renamed to R_BC();
5. calculate_summary_distances_of_cells_to_borders() - one column
in the returned data frame has name change;
6. defined_strcture() - parameter name_of_immune_cells renamed to
cell_types_of_interest;
- identify_bordering_cells() can return the number of clusters of the
specified cell type.
BUG FIXES
- Fixed the calculation of the normalised cross-K AUC in
AUC_of_cross_funcion().
- identify_neighborhood() returns an SPE object instead of sending
ERROR when the cells of interest do no form any clusters.
Changes in version 0.99.4
BUG FIXES
- Fixed internal function bind_info() to avoid duplicated columns.
Changes in version 0.99.3
BUG FIXES
- Added checks for dimensionality_reduction_plot() to return error
when the cell count in an image is too low.
Changes in version 0.99.2
SIGNIFICANT USER-VISIBLE CHANGES
- Added perplexity parameter to dimensionality_reduction_plot().
- Added plot_final_border parameter to identify_bordering_cells().
BUG FIXES
- image_splitter() returns NULL for the sub-images that do not contain
any cells.
- calculate_spatial_autocorrelation()
- identify_bordering_cells() only plots the bordering cells when
bordering cells are detected and the user chooses to plot them.
- The calculation of one cell type to All_cells_in_the_structure in
calculate_proportions_of_cells_in_structure() was incorrect. Now
fixed.
Changes in version 0.99.1
SIGNIFICANT USER-VISIBLE CHANGES
- Updated the main object class from SingleCellExperiment to
SpatialExperiment.
- Replace "sce" with "spe" in function names.
- Reading data by format_image_to_spe() using general format is
recommended. For other data platforms, format_inform_to_spe(),
format_halo_to_spe(), format_codex_to_spe(), and
format_cellprofiler_to_spe() are also available.
- Deleted "Visium" format option in format_image_to_spe().
- image_splitter() now returns a list of spe objects.
Changes in version 0.99.0
Version submitted to Nature Communications. Access the release v0.99.0
here.
SIGNIFICANT USER-VISIBLE CHANGES
- Systematically renamed parameter names. For example, the parameter
for selecting a column of interest is changed from "column" to
"feature_colname".
- Added a function entropy_gradient_aggregated() for calculating cell
colocalisation.
- Added a function dimensionality_reduction_plot() for visualisation
and quality control.
- Added examples to documentation. The example datasets were simulated
by spaSim.
Changes in version 0.4
SIGNIFICANT USER-VISIBLE CHANGES
- Added a new parameter "column" to most functions to select a column
of interest.
- Users can define cell types based on certain marker combinations or
cell properties.
- Improved tumour border detection (identify_bordering_cells()).
- Added functions to define tumour structure and characterise immune
populations in different tissue regions.
- Added functions to characterise spatial heterogeneity.
- Added functions for cell colocalisation, including mixing score,
normalised mixing score, cross K function.
Changes in version 0.2
- Version for the paper submitted to BioRxiv.