ewce_plot functionality more clear - will no fail if
make_dendro=TRUE and ggdendro is not installed or CTD is not provided
rather than issuing a warning.check_bootstrap_args:
drop_uninformative_genes:
check_sce() function.bootstrap_enrichment_test:
standardise_sct_data=, standardise_hits=:
let users have more control over data standardisation steps.check_ewce_genelist_inputs: updated accordingly.store_gene_data to avoid hitting memory limits.ewce_plot() - Dendrogram not reordering cell types in plot
make_dendro = TRUE for ewce_plot().generate_bootstrap_plots
gene_data object whenever possible.generate_bootstrap_plots
check_ewce_genelist_inputs() call.drop_uninformative_genes
check_species:
sctSpecies_origin_defaultstandardise_ctd:
sctSpecies_origin, which will be added to the metadata.standardise_ctd:
NULL.check_ewce_genelist_inputs/bootstrap_enrichment_test
sctSpecies_origin lets users clarify that their data originally
came from mouse even when it is currently formatted as human orthologs.
This is necessary for creating the appropriate background gene lists.grDevices as dep entirely.fix_celltype_names
make_unique to make this function easily usable for vectors
where the same celltype appears multiple times.bootstrap_enrichment_test
generate_bootstrap_plots.
now stored as a list element named gene_data in data.table format.generate_bootstrap_plots
gene_data.
It will also tell you which of these options it's using.ggsave instead if grDevices.adj_pval_threshsavePath arg to the more accurate save_dir. Expose appending BootstrapPlots to the user within the argument.generate_bootstrap_plots_for_transcriptome
savePath arg to the more accurate save_dir. Expose appending BootstrapPlots to the user within the argument.ggsave instead if grDevices.hits + hitGenes arg all to hits.drop_uninformative_genes / generate_celltype_data
verbose arg to matrix formatting functions.generate_controlled_bootstrap_geneset
combinedGenes arg as it was not being used anywhere within.check_args_for_bootstrap_plot_generation
ttSpecies, sctSpeciesorthogene databases improving.rworkflows GHA.
rworkflows::use_badges to README.Rmd.ewceData (>=1.7.1) is now required, due to a fix made only
in the development version of rtracklayer.cowplot dependency.%>% with |>calc_quantiles:
filter_variance_quantilesstats::ecdf vs. dplyr::ntile methods.EWCE as it's not longer used anywhere.bin_columns_into_quantiles:
matrixIn --> vec to reflect what the function actually does.filter_variance_quantiles:
bin_columns_into_quantiles instead of calc_quantiles
to be consistent with how quantiles are handled in the rest of EWCE.stats::quantile).ewce_plot:
gridArrange/cowplot to patchwork.ewceData). These functions have the parameter localhub added to
control this.standardise_ctd:
filter_ctd_genes
get_ctd_matrix_names: New function to get a list of all data matrices in CTD.check_ewce_genelist_inputs:
standardise_ctd:
check_species()as_sparse=TRUE.fix_celltype_names:
genelistSpecies now passed to prepare_genesize_control_network in
bootstrap_enrichment_test meaning gene list species will be inferred from user
input.drop_uninformative_genes:
dge_method, dge_test, min_variance_decilemerged_ctd: Actually merge the CTDs into one when as_SCE=FALSE.sprintf("%s/MRK_List2.rpt", tempdir())) to be more compatible with Windows.orthogene, so going through and making sure everything still works / is able to take advantage of new features (e.g. separation of non121_strategy and agg_func args, many:many mapping):
filter_nonorthologs: Pass up args from orthogene::convert_orthologs.generate_celltype_data: @inheritDotParamsmethod argument from orthogene::create_background and orthogene::convert_orthologs is now passed up as an argument to EWCE functions to give users more control. "homologene" chosen as default for all functions. "homologene" has fewer species than "orthogene" but doesnt need to import data from the web. It also has more 1:1 mouse:human orthologs.bin_specificity_into_quantiles to set specificity matrix
name produced.try({}) and error=TRUE to avoid "polygon edge not found" error in vignettes.orthogene.standardise_ctd.DelayedArray object class.plot_ctd).EWCE::example_bootstrap_results().BiocCheck, and rebuild/deploy pkgdown site.drop_uninformative_genesgenerate_celltype-databootstrap_enrichment_testewceData files have to be downloaded):
test-DelayedArraytest-merge_scetest-get_celltype_tabletest-list_speciestest-run_DGEtest-check_percent_hitsis_32bit() to all tests to ensure they don't
get run twice on Windows OS.check-bioc-docker.yml: Runs CRAN/Bioc checks, rebuilds and pushes pkgdown website, runs and uploads test coverage report,dockerhub.yml: Builds Bioconductor Docker container with EWCE installed, runs CRAN checks and (if checks are successful) pushes container to neurogenomicslab DockerHub.docs folder, as the documentation website comes from the
gh-pages branch now, and is automatically built by GHA workflow
after each push to main branch.fix_celltype_names to help with standardising
celltype names in alignment with standardise_ctd.generate_bootstrap_plots_for_transcriptome: Now supports any species
(not just mouse or human).
tt) into output_species gene symbols.ewce_expression_data via new example_transcriptome_results function.@return documentation for internal functions.test_ ==> test-drop_uninformative_genes for now until we run
benchmarking to see how each affects the EWCE results.bootstrap_plots function internal.orthogene improve within- and across-species
gene mappings in extended vignette.standardise_ctd output:
input_species and output_speciesEWCE, orthogene, and homologenedrop_uninformative_genes() has been expanded to allow the utilisation of differential expression approachesDeprecated & Defunct
check.ewce.genelist.inputscell.list.distbootstrap.enrichment.testbin.specificity.into.quantilesbin.columns.into.quantilesadd.res.to.merging.listprepare.genesize.control.networkprep.dendroget.celltype.tablecalculate.specificity.for.levelcalculate.meanexp.for.levelgenerate.celltype.datagenerate.bootstrap.plotsgenerate.bootstrap.plots.for.transcriptomefix.bad.mgi.symbolsfix.bad.hgnc.symbolsfilter.genes.without.1to1.homologewce.plotcells.in.ctddrop.uninformative.genes