- Added a catch and correction for special characters present in cell type names
- Removed ssvd dependency and sparse tucker implementation
- Updated readme to include instructions for installing other dependencies
- Fixed error in LR analysis caused by genes with 0 expression
- New function for users to easily get gene significance p-values for a factor
- Additional url for new LR analysis tutorial in the README
- Enabled
compute_LR_interact
to run when data contains alternate gene symbols - Included warning/catch in
plot_loadings_annot
if no genes are significant - Fixed bug in
determine_ranks_tucker
with variance scaling in wrong order - Updated README figure
- Added
count_word()
function from simplifyEnrichment package as the remoevd it - Included necessary suggests packages for
count_word()
- Updated README to include our preprint citation and new figure
- Optional parameters
min_gs_size
andmax_gs_size
torun_gsea_one_factor()
- New function project_new_data() to project a multicellular pattern onto new data
- Warning if cell-level metadata is included as donor-level metadata
- Check that cell type names in
ctypes_use
are spelled correctly
- Checks for empty batches before running ComBat batch correction
- Fixed label issue with LR output when only have associations with 1 factor
- Option to use custom list of genes in the tensor
- Check for matching cell names between count matrix and metadata
- Removed
optimize_var_scale_power()
and helpers as they're no longer used - Removed
compare_factors()
and helpers as they're no longer used
- Fixed
anova
vargenes method for genes with zero expression across donors - Updated
run_jackstraw()
andget_min_sig_genes()
to work with new rotations - Fixed bug in
plot_subclust_associations()
- Enabled storage of results from
get_subtype_prop_associations()
- Updated README to no longer require dev version of ComplexHeatmap
- Updated DESCRIPTION authors and description