NEWS

---

goatea 2.1.0

• polishing by CRAN standards for publishing
• created NEWS.md from changelog

goatea 0.2.5

• renamed package to lowercase goatea
• fixed dplyr %>% pipelines with .data$undefined_var_names
• fixed subgraphing/filtering by highlighting clusters/features
  • noted that highlighted but isolated (disconnected) nodes are not graphed
  • noted that clustering of the selected proteins is kept, unless if the proteins selected all belong to the same cluster
• fixed object g not found error by getting it from reactive value
• added save to R base folder and load loaded_taxid_genesets.Rdata functionality
• added showNotification for ppi subgraph and subheatmaps
• fixed alignment of signif and signif_overlap in gene*geneset heatmap
• if GO_BP in names, default for 'Show enrichment source'
• transforming toupper(genelist$symbol) when mapping in order to format mapIds ALIAS
  • warning on %mapping not converted to ENTREZIDs
• noted to download org.Xx.eg.db manually!
• fixed reloading data by resetting earlier loaded data, requiring to load all wanted data in one upload
• set contentType to NULL on downloadHandlers to use global format parameter for tables to auto use .csv or .xlsx
• kept only explicit ::fromList() calls instead of importFrom UpSetR or upsetjs to remove message for BiocCheck
• added showNotifications when gene selection buttons clicked

goatea 0.2.0

• added priority of selected genes to gene*set heatmap
• fixed setting names to updateSelectInputs and updating names in gene_overview
• set all :: and/or importFrom for R CMD check/BiocChecks
• for gene*effectsize heatmap: fixed setting as many genes in annotation as in matrix

goatea 0.1.19

• visnetwork ppigraphs hover tooltip background and labels colored via JavaScript such that it is readable (www/visnetwork_hover_tooltip_style.js)
• genes*genesets heatmap: showing top genes based on ordering by highest ncount in terms and then highest effectsize
• fixed interactive overlap: upsetjs::chartTheme()
• added max gene filters to gene*geneset heatmap
• ppi
  • fixed hover node labels with all info: added toupper() to match gene and protein symbols
  • fixed coloring by L2FC: added mutate(toupper()) to match gene and protein symbols
  • enabled entering gene/cluster to highlight or center on
  • using set_base_folder() for PPI STRINGdb downloads
  • added functionality to subgraph/filter/delete highlighted nodes/edges by numeric feature
• added modal dialogues per tab
• set_significant_N_genes: changed code order to not display warning when keeping max N genes

goatea 0.1.18

• ppi subgraph: keep original graph clustering
• added download button for vanilla EnhancedVolcano plot
• Set and calculate gene_overview when setting significant genes
  • needed for ppi metadata coloring
  • moved download gene overview metadata button to 'load data' tab below set names
• added interactive gene*L2FC heatmap figure for all loaded samples and user selected genes

goatea 0.1.17

• set term n=50 and gene n=100 defaults for heatmap
• kevinblighe/EnhancedVolcano answered issues 77/114 and Biostars 421750 by making minimal reproducible example of shiny + EnhancedVolcano + plotly for other users
• added interactive EnhancedVolcano plot via ploty::ggplotly()

goatea 0.1.16

• added interactive UpSet plot (removed Venn completely)
• added overlap gene selection

goatea 0.1.15

• started changelog
• added to load from GMT file: goat::load_genesets_gmtfile(), label = "GMT")
  • checked that genesets cannot be loaded without setting 'source' column, for downstream processing
• fixed run_geneset_enrichment to actually use the given parameters
• added set_significant_N_genes() function

Future steps

• sort heatmap by zscore (up/down regulation)
• genefsi_icheatmap_action functionalize for selecting/removing/resetting genes (selection)
• plot: similarity_heatmap(genesets/termsets) (see R/testing/_plot_similarity_heatmap.R)
• volcano user label axes
• interactive termtree
• interactive splitdot
• selected genes tab
• selected terms tab (if used for plotting)
• select genes by frequency of them popping up in terms
• add genes to selection via input file (& custom genes selection tab?)
• ppigraph: add terms as different node shape, edges to those could be dashes to differentiate
• network graph
  • terms overlap by ngenes graph (connection shows overlap size, node size = ngenes)
    • may be able to add genes to this graph?
• make menu_tab for colourpicker::colourInput to have user defined colors for plots
• heatmap: significant genes, saturated (alpha) hue, otherwise less saturated
• check heatmap y axis label lengths (going out of plot ?)