use load_pack more consistently; started cp-utils

R/bio/bioinfo_commons.R

@@ -2,7 +2,7 @@
 #library(Biostrings, quietly=T)
 #require_auto(BiocGenerics x
 
-#loadpack(Biostrings)
+#load_pack(Biostrings)
 
 ## install Biostrings if not yet there
 install_package("Biostrings")

R/bio/cp_utils.R

+install_package(clusterProfiler)
+
+
+## todo move to diffex commons
+guess_cp_species <- function(ensIds){
+    an_id <-ensIds[1]
+
+    if(str_detect(an_id, "ENSG")){
+        return("human")
+    }else if(str_detect(an_id, "ENSMUSG")){
+        return("mouse")
+    }else if(str_detect(an_id, "ENSDARG")){
+        return("zebrafish")
+    }else if(str_detect(an_id, "FBgn")){
+        return("fly")
+    }else{
+        stop(paste("could not clusterProfiler species name from ", an_id))
+    }
+}
+
+guess_anno_db <- function(ensIds){
+    an_id <-ensIds[1]
+
+    if(str_detect(an_id, "ENSG")){
+        return("org.Hs.eg.db")
+    }else if(str_detect(an_id, "ENSMUSG")){
+        return("org.Mm.eg.db")
+    }else if(str_detect(an_id, "ENSDARG")){
+        return("org.Dr.eg.db")
+    }else if(str_detect(an_id, "FBgn")){
+        return("org.Dm.eg.db")
+    }else{
+        stop(paste("could not anno db mart from ", an_id))
+    }
+}
+
+#source("http://bioconductor.org/biocLite.R")
+#biocLite("org.Mm.eg.db")
+#biocLite("org.Hs.eg.db")
+#biocLite("org.Dr.eg.db")
+#biocLite("org.Dm.eg.db")
+#biocLite("KEGG.db")
+
+
+
+load_pack(ReactomePA)
+
+cp_test <- function(geneIds){
+    # DEBUG geneIds <- glMapped %>% filter(cluster %in% c("cluster_9")) %$% entrez_gene_id %>% as.integer
+    # DEBUG geneIds <- head(glMapped,30)$entrez_gene_id
+    #    geneIds=.$entrez_gene_id
+
+    if(length(geneIds)>1500){
+        geneIds <- sample(geneIds) %>% head(1500)
+    }
+
+    echo("testing", length(geneIds), " genes for enrichment")
+
+    #    PANTHER10_ontology <- read.delim("http://data.pantherdb.org/PANTHER10.0/ontology/Protein_Class_7.0")
+
+    #    browser()
+    #    pantherResults <-     enricher(gene = geneIds, organism = cpSpecies, qvalueCutoff = qCutoff, readable = TRUE, TERM2GENE = PANTHER10_ontology) %>% summary()
+    keggResults <-        clusterProfiler::enrichKEGG(gene = geneIds, organism = cpSpecies, qvalueCutoff = qCutoff, use_internal_data=T) %>% summary()
+    reactomeResults <-      enrichPathway(gene = geneIds, organism = cpSpecies, qvalueCutoff = qCutoff) %>% summary()
+    goResultsCC <-          clusterProfiler::enrichGO(gene = geneIds, OrgDb = annoDb, qvalueCutoff = qCutoff, ont = "CC") %>% summary()
+    goResultsMF <-          clusterProfiler::enrichGO(gene = geneIds, OrgDb = annoDb, qvalueCutoff = qCutoff, ont = "MF") %>% summary()
+    goResultsBP <-          clusterProfiler::enrichGO(gene = geneIds, OrgDb = annoDb, qvalueCutoff = qCutoff, ont = "BP") %>% summary()
+
+    #cp-bug: if no pathways are enriched odd strucuture is retured ##todo file issue
+    if(!("data.frame" %in% class(keggResults))) keggResults <- filter(goResultsBP, Description="foobar")
+
+    enrResults <- bind_rows(
+    mutate(keggResults, ontology="kegg"),
+    mutate(reactomeResults, ontology="reactome"),
+    mutate(goResultsBP, ontology="go_bp"),
+    mutate(goResultsMF, ontology="go_mf"),
+    mutate(goResultsCC, ontology="go_cc")
+    )
+    enrResults
+    #    echo("numResults", nrow(enrResults))
+}
+
+
+## example
+if(F){
+
+someGenes = c("") ## todo continue
+
+cpSpecies <- guess_cp_species(someGenes)
+annoDb <- guess_anno_db(someGenes) # e.g. "org.Hs.eg.db"
+
+
+enrResults <- cp_test(someGenes)
+## or for grouped data
+#enrResults <-  quote(glMapped %>% do(cp_test(.$entrez_gene_id))) %>% cache_it(paste0("enrdata_", digest(glMapped)))
+
+## test custom ontology
+#enrichrFile_TFchip = "/projects/bioinfo/holger/data/enrichr_datasets/ENCODE_TF_ChIP-seq_2015.txt"
+#tfChipOnt <- convert_enrichr_cp(enrichrFile_TFchip)
+
+}
\ No newline at end of file

R/core_commons.R

@@ -31,7 +31,7 @@ options(gsubfn.engine = "R")
 ## automatic package installation
 
 require_auto <-  function(x){
-    warning("require_auto is deprecated, use loadpack() instead")
+    warning("require_auto is deprecated, use load_pack() instead")
 
     x <- as.character(substitute(x))
 
@@ -75,7 +75,7 @@ load_pack <-  function(x, warn_conflicts=T){
 
    install_package(x)
 
-    ## load it using a library function so that loadpack errors if package is still not ins
+    ## load it using a library function so that load_pack errors if package is still not ins
     eval(parse(text=paste("library(", x, ",  quietly=T, warn.conflicts=", warn_conflicts, ")", sep="")))
 }
 
@@ -86,7 +86,7 @@ loadpack <-  function(x, warn_conflicts=T){
 
    install_package(x)
 
-    ## load it using a library function so that loadpack errors if package is still not ins
+    ## load it using a library function so that load_pack errors if package is still not ins
     eval(parse(text=paste("library(", x, ",  quietly=T, warn.conflicts=", warn_conflicts, ")", sep="")))
 }
 
@@ -100,34 +100,34 @@ check_version = function(pkg_name, min_version) {
 ########################################################################################################################
 ## load core packages
 
-loadpack(plyr)
-loadpack(stringr)
-loadpack(reshape2)
-#loadpack(reshape2, quietly=T, warn_conflicts=F)
+load_pack(plyr)
+load_pack(stringr)
+load_pack(reshape2)
+#load_pack(reshape2, quietly=T, warn_conflicts=F)
 
 ## load on purpose after plyr
-loadpack(dplyr, warn_conflicts=F)
-loadpack(magrittr, warn_conflicts=F)
-loadpack(tidyr, warn_conflicts=F)
+load_pack(dplyr, warn_conflicts=F)
+load_pack(magrittr, warn_conflicts=F)
+load_pack(tidyr, warn_conflicts=F)
 
 ## needed for caching
-loadpack(digest)
-loadpack(readr)
-suppressWarnings(loadpack(readxl)) ## supress differring build number
-#loadpack(readxl) ## supress differring build number
+load_pack(digest)
+load_pack(readr)
+suppressWarnings(load_pack(readxl)) ## supress differring build number
+#load_pack(readxl) ## supress differring build number
 
 
 ## common plotting requirements since they are omnipresent
-loadpack(ggplot2)
-loadpack(scales, warn_conflicts=F)
-loadpack(grid)
+load_pack(ggplot2)
+load_pack(scales, warn_conflicts=F)
+load_pack(grid)
 
 ## for table exploration without using Rstudio
-loadpack(DT)
+load_pack(DT)
 
 
 ## moved into datatable_commons because replaced almost everywhere with dplyr
-#loadpack(data.table)
+#load_pack(data.table)
 
 
 
@@ -461,6 +461,6 @@ assert <- function (expr, error) {
 }
 
 ### table rendering
-table_browser <- function(df){
-    datatable(df, filter = "bottom", extensions = 'Buttons', options = list( dom = 'Bfrtip', buttons = c('copy', 'csv', 'excel')))
+table_browser <- function(df, ...){
+    datatable(df, filter = "bottom", extensions = 'Buttons', options = list( dom = 'Bfrtip', buttons = c('copy', 'csv', 'excel')), ...)
 }

R/ggplot/ggheatmap.R

@@ -2,8 +2,8 @@
 #library(ggplot2)
 #library(reshape2)
 
-require_auto(ggdendro)
-require_auto(grid)
+load_pack(ggdendro)
+load_pack(grid)
 
 ## Adopted from  http://cwcode.wordpress.com/2013/01/30/ggheatmap-version-2/