Title: BACA: Bubble chArt to Compare Annotations by using DAVID
Author: “Vittorio Fortino and Dario Greco”
In some cases you have different, large lists of genes and you would like to perform several enrichment analysis and graphically compare the corresponding results. This can be done using the R package BACA.
BACA provides three R functions: DAVIDSearch, BBplot and Jplot.
DAVIDSearch is a user-friendly R function that wraps the code necessary to use the different types of functional annotations available in DAVID knowledgebase. It uses the R package RDAVIDWebService to query DAVID and wrap the results into R objects namely DAVIDFunctionalAnnotationChart. For more details please visit http://david.abcc.ncifcrf.gov/content.jsp?file=WS.html.
DAVIDSearch() accepts in input different lists of up-/down-regulated genes, the email of a given registered DAVID users, the type of gene ID, the type of list (Gene or Background), the EASE enrichment score, the functional type of annotation and the specie to use (see DAVID for more details). After querying DAVID web services (http://david.abcc.ncifcrf.gov/webservice/register.htm), it outputs a list of DAVIDFunctionalAnnotationChart objects, one for each specified list of genes.
DAVIDsearch
BBplot is an R function that uses the DAVIDFunctionalAnnotationChart objects to build a chart that graphically compares functional annotations found by DAVID. This chart shows a grid where each row represents an enriched annotation and each column corresponds to a condition/treatment where that annotation was highlighted. While, each cell reports a bubble indicating the number of genes enriching the corresplonding annotation and the state of these genes in terms of down- and up-regulation (default setting: green = “down” - red = “up”). BBplot function can work with any kind of gene list of interest and not only with up/down regulated gene lists. Moreover, it allows users to compare the enrichment of differential genes for different subgroups.
BBplot
BACA uses external packages and assumes that they are installed. Packages to install and load before to use BACA: RDAVIDWebService and ggplot2.
Examples.
library(RDAVIDWebService)
library(ggplot2)
library(BACA)
The example will be carried out using artificial gene lists: gene.lists.ex. This data contains artifical up- and down-regulated gene lists corresponding to two time points of two different experimental conditions.
data(gene.lists.ex)
The str() function can be used to check the content of gene.lists.ex.
str(gene.lists.ex)
## List of 8
## $ dn_cond.1_12h: chr [1:66] "12876" "67717" "433904" "72185" ...
## $ up_cond.1_12h: chr [1:369] "14537" "18971" "74424" "59126" ...
## $ dn_cond.1_24h: chr [1:731] "17289" "12808" "59049" "99586" ...
## $ up_cond.1_24h: chr [1:1157] "17159" "637515" "14113" "18226" ...
## $ dn_cond.2_12h: chr [1:756] "23873" "70601" "74211" "621542" ...
## $ up_cond.2_12h: chr [1:1170] "72026" "327958" "104009" "21841" ...
## $ dn_cond.2_24h: chr [1:1537] "12616" "18769" "20112" "66343" ...
## $ up_cond.2_24h: chr [1:1569] "664987" "26447" "114606" "67561" ...
The gene.lists.ex contains eight character vectors:
1) Condition #1 - time point 12h (list of down-regulated genes).
2) Condition #1 - time point 12h (list of up-regulated genes).
3) Condition #1 - time point 24h (list of down-regulated genes).
4) Condition #1 - time point 24h (list of up-regulated genes).
5) Condition #2 - time point 12h (list of down-regulated genes).
6) Condition #2 - time point 12h (list of up-regulated genes).
7) Condition #2 - time point 24h (list of down-regulated genes).
8) Condition #2 - time point 24h (list of up-regulated genes).
After loading the data, use DAVIDsearch() to query the DAVID knowledge base. It requires two main inputs: (1) the lists of up-/down-regulated gene sets and (2) the email of a given registered DAVID users (make sure you are registered at http://david.abcc.ncifcrf.gov/webservice/register.htm). Then other optional inputs can be specified.
# result.kegg <- DAVIDsearch(gene.lists.ex, david.user = "###########",
# idType="ENTREZ_GENE_ID", annotation="KEGG_PATHWAY")
Note, DAVID will find the species associated to the submitted gene lists and DAVIDsearch function will print out the corresponding names. Therefore, if the user wants to select one species name, it can re-use DAVIDsearch() function indicating as “species” a different, found pecies name.
Examples of annotation that users can specify: “BIOCARTA”, “ENSEMBL_GENE_ID”, “ENTREZ_GENE_ID”, “GOTERM_CC_ALL”, “GOTERM_BP_ALL”, “GOTERM_BP_ALL”, “KEGG_PATHWAY”, “INTERPRO”, etc. While, examples of “idType” that users can specify: “GENBANK_ACCESSION”, “ENSEMBL_GENE_ID”, “PFAM_ID”, etc.
Please, use the functions getAllAnnotationCategoryNames() and getIdTypes(david.obj) of the R package RDAVIDWebService to prin out all the available annotation category names and DAVID idTypes.
## [1] "BBID"
## [2] "BIND"
## [3] "BIOCARTA"
## [4] "BLOCKS"
## [5] "CGAP_EST_QUARTILE"
## [6] "CGAP_SAGE_QUARTILE"
## [7] "CHROMOSOME"
## [8] "COG_NAME"
## [9] "COG_ONTOLOGY"
## [10] "CYTOBAND"
## [11] "DIP"
## [12] "EC_NUMBER"
## [13] "ENSEMBL_GENE_ID"
## [14] "ENTREZ_GENE_ID"
## [15] "ENTREZ_GENE_SUMMARY"
## [16] "GENETIC_ASSOCIATION_DB_DISEASE"
## [17] "GENERIF_SUMMARY"
## [18] "GNF_U133A_QUARTILE"
## [19] "GENETIC_ASSOCIATION_DB_DISEASE_CLASS"
## [20] "GOTERM_BP_2"
## [21] "GOTERM_BP_1"
## [22] "GOTERM_BP_4"
## [23] "GOTERM_BP_3"
## [24] "GOTERM_BP_FAT"
## [25] "GOTERM_BP_5"
## [26] "GOTERM_CC_1"
## [27] "GOTERM_BP_ALL"
## [28] "GOTERM_CC_3"
## [29] "GOTERM_CC_2"
## [30] "GOTERM_CC_5"
## [31] "GOTERM_CC_4"
## [32] "GOTERM_MF_1"
## [33] "GOTERM_MF_2"
## [34] "GOTERM_CC_FAT"
## [35] "GOTERM_CC_ALL"
## [36] "GOTERM_MF_5"
## [37] "GOTERM_MF_FAT"
## [38] "GOTERM_MF_3"
## [39] "GOTERM_MF_4"
## [40] "HIV_INTERACTION_CATEGORY"
## [41] "HIV_INTERACTION_PUBMED_ID"
## [42] "GOTERM_MF_ALL"
## [43] "HIV_INTERACTION"
## [44] "KEGG_PATHWAY"
## [45] "HOMOLOGOUS_GENE"
## [46] "INTERPRO"
## [47] "OFFICIAL_GENE_SYMBOL"
## [48] "NCICB_CAPATHWAY_INTERACTION"
## [49] "MINT"
## [50] "PANTHER_MF_ALL"
## [51] "PANTHER_FAMILY"
## [52] "PANTHER_BP_ALL"
## [53] "OMIM_DISEASE"
## [54] "PFAM"
## [55] "PANTHER_SUBFAMILY"
## [56] "PANTHER_PATHWAY"
## [57] "PIR_SUPERFAMILY"
## [58] "PIR_SUMMARY"
## [59] "PIR_SEQ_FEATURE"
## [60] "PROSITE"
## [61] "PUBMED_ID"
## [62] "REACTOME_INTERACTION"
## [63] "REACTOME_PATHWAY"
## [64] "PIR_TISSUE_SPECIFICITY"
## [65] "PRINTS"
## [66] "PRODOM"
## [67] "PROFILE"
## [68] "SMART"
## [69] "SP_COMMENT"
## [70] "SP_COMMENT_TYPE"
## [71] "SP_PIR_KEYWORDS"
## [72] "SCOP_CLASS"
## [73] "SCOP_FAMILY"
## [74] "SCOP_FOLD"
## [75] "SCOP_SUPERFAMILY"
## [76] "UP_SEQ_FEATURE"
## [77] "UNIGENE_EST_QUARTILE"
## [78] "ZFIN_ANATOMY"
## [79] "UP_TISSUE"
## [80] "TIGRFAMS"
## [81] "SSF"
## [82] "UCSC_TFBS"
## [1] "AFFYMETRIX_3PRIME_IVT_ID" "AFFYMETRIX_EXON_GENE_ID"
## [3] "AFFYMETRIX_SNP_ID" "AGILENT_CHIP_ID"
## [5] "AGILENT_ID" "AGILENT_OLIGO_ID"
## [7] "ENSEMBL_GENE_ID" "ENSEMBL_TRANSCRIPT_ID"
## [9] "ENTREZ_GENE_ID" "FLYBASE_GENE_ID"
## [11] "FLYBASE_TRANSCRIPT_ID" "GENBANK_ACCESSION"
## [13] "GENOMIC_GI_ACCESSION" "GENPEPT_ACCESSION"
## [15] "ILLUMINA_ID" "IPI_ID"
## [17] "MGI_ID" "PFAM_ID"
## [19] "PIR_ID" "PROTEIN_GI_ACCESSION"
## [21] "REFSEQ_GENOMIC" "REFSEQ_MRNA"
## [23] "REFSEQ_PROTEIN" "REFSEQ_RNA"
## [25] "RGD_ID" "SGD_ID"
## [27] "TAIR_ID" "UCSC_GENE_ID"
## [29] "UNIGENE" "UNIPROT_ACCESSION"
## [31] "UNIPROT_ID" "UNIREF100_ID"
## [33] "WORMBASE_GENE_ID" "WORMPEP_ID"
## [35] "ZFIN_ID"
DAVIDSearch outputs a list of DAVIDFunctionalAnnotationChart objects, one for each specified list of genes. The DAVIDFunctionalAnnotationChart class is defined into the R package http://www.bioconductor.org/packages/release/bioc/html/RDAVIDWebService.html, it represents the output of “Functional Annotation Chart” of DAVID.
Note, query to DAVID webserver can take a long time to execute. Therefore, BACA package provides the DAVIDFunctionalAnnotationChart objects obtained querying DAVID with the artificial gene lists in gene.lists.ex.
data(result.kegg)
After using DAVIDsearch, build the bubble plot using the BBplot() function.
bbplot.kegg <- BBplot(result.kegg, max.pval = 0.05, min.ngenes = 10,
name.com = c("C1_12h","C1_24h","C2_12h","C2_24h"),
labels = c("down", "up"), colors = c("#009E73", "red"),
title = "BBplot - KEGG")
bbplot.kegg
Save the plot as a TIFF image using the ggsave function in the R package ggplot2.
ggsave("KEGG_terms.tiff", width=6, height=4, scale=2, dpi=200)
Finally, use the Jplot() function to build/plot pairwise comparisons between functional annotation charts.
jplot.kegg <- Jplot(result.kegg[[4]], result.kegg[[2]],
max.pval = 0.05, min.ngenes = 10,
print.term="name")
jplot.kegg
