3.1 Sample 1, start from given peptides positions

3.2 Step 1, fetch sequences

You should have interesting peptides position info and the identifiers for fetching sequences via biomaRt.

library(dagLogo)
library(biomaRt)
try({##just in case biomaRt server does not response
    mart <- useMart("ensembl", "dmelanogaster_gene_ensembl")
    dat <- read.csv(system.file("extdata", "dagLogoTestData.csv", 
                                package="dagLogo"))
    dat <- dat[1:5,] ##subset to speed sample
    dat
    seq <- fetchSequence(as.character(dat$entrez_geneid), 
                         anchorPos=as.character(dat$NCBI_site), 
                         mart=mart, 
                         upstreamOffset=7, 
                         downstreamOffset=7)
    head(seq@peptides)
})

##      [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13]
## [1,] "G"  "I"  "A"  "S"  "E"  "A"  "Q"  "K"  "Y"  "Q"   "A"   "K"   "I"  
## [2,] "A"  "S"  "K"  "V"  "A"  "L"  "S"  "K"  "F"  "D"   "S"   "D"   "V"  
## [3,] "Q"  "F"  "I"  "S"  "S"  "G"  "L"  "K"  "K"  "V"   "A"   "V"   "P"  
## [4,] "G"  "R"  "C"  "A"  "S"  "I"  "A"  "K"  "D"  "A"   "M"   "S"   "H"  
## [5,] "G"  "I"  "S"  "E"  "V"  "F"  "D"  "K"  "F"  "G"   "G"   "T"   "V"  
##      [,14] [,15]
## [1,] "L"   "S"  
## [2,] "Y"   "L"  
## [3,] "S"   "T"  
## [4,] "G"   "L"  
## [5,] "L"   "A"

Sometimes you don’t have the exactly postions. You have only the interesting peptides and the identifiers.

try({
    seq <- fetchSequence(as.character(dat$entrez_geneid), 
                         anchorAA="*",
                         anchorPos=as.character(dat$peptide), 
                         mart=mart, 
                         upstreamOffset=7, 
                         downstreamOffset=7)
    head(seq@peptides)
})

##      [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13]
## [1,] "G"  "I"  "A"  "S"  "E"  "A"  "Q"  "K"  "Y"  "Q"   "A"   "K"   "I"  
## [2,] "A"  "S"  "K"  "V"  "A"  "L"  "S"  "K"  "F"  "D"   "S"   "D"   "V"  
## [3,] "Q"  "F"  "I"  "S"  "S"  "G"  "L"  "K"  "K"  "V"   "A"   "V"   "P"  
## [4,] "G"  "R"  "C"  "A"  "S"  "I"  "A"  "K"  "D"  "A"   "M"   "S"   "H"  
## [5,] "G"  "I"  "S"  "E"  "V"  "F"  "D"  "K"  "F"  "G"   "G"   "T"   "V"  
##      [,14] [,15]
## [1,] "L"   "S"  
## [2,] "Y"   "L"  
## [3,] "S"   "T"  
## [4,] "G"   "L"  
## [5,] "L"   "A"

In above sample, anchorAA is represented by asterisk. In following sample, anchorAA is represented by lower case of amino acid.

if(interactive()){
    dat <- read.csv(system.file("extdata", "peptides4dagLogo.csv",
                package="dagLogo"))

    tail(dat)
    mart <- useMart("ensembl", "hsapiens_gene_ensembl")
    seq <- fetchSequence(toupper(as.character(dat$symbol)), 
                         type="hgnc_symbol",
                         anchorAA="s",
                         anchorPos=as.character(dat$peptides), 
                         mart=mart, 
                         upstreamOffset=7, 
                         downstreamOffset=7)
    head(seq@peptides)
}

Sometimes you may already have the aligned peptides sequences in hand. You will use formatSequence function to prepare an object of dagPeptides for further testing. To use formatSequence, you need prepare the proteome by prepareProteome function.

dat <- unlist(read.delim(system.file("extdata", "grB.txt", package="dagLogo"), 
                         header=F, as.is=TRUE))
head(dat)

##                              V11                              V12 
## "GHISVKEPTPSIASDISLPIATQELRQRLR" "EREMFDKASLKLGLDKAVLQSMSGRENATN" 
##                              V13                              V14 
## "XXXXXXMSDIVVVTDLIAVGLKRGSDELLS" "GQDQEEEEIEDILMDTEEELMRAEDTEQLK" 
##                              V15                              V16 
## "ESYATDNEKMTSTPETLLEEIEAKNRELIA" "VENKERTLKRLLLQDQENSLQDNRTSSDSP"

##prepare proteome from a fasta file
proteome <- prepareProteome(fasta=system.file("extdata", 
                                              "HUMAN.fasta",
                                              package="dagLogo"))
##prepare object of dagPeptides
seq <- formatSequence(seq=dat, proteome=proteome, 
                      upstreamOffset=14, downstreamOffset=15)

3.3 Step 2, build background model

Once you have an object of dagPeptides in hand, you can start to build background model for DAG test. The background could be random subsequence of whole proteome or your inputs. If the background was built from whole proteome or proteome without your inputs, an object of Proteome is required.

To prepare a proteome, there are two methods, from a fasta file or from UniProt webservice. Last example shows how to prepare proteome from a fasta file. Here we show how to prepare proteome via UniProt webservice.

if(interactive()){
    library(UniProt.ws)
    UniProt.ws <- UniProt.ws(taxId=9606)
    proteome <- prepareProteome(UniProt.ws=UniProt.ws)
}

Then the proteome can be used for background model building.

bg <- buildBackgroundModel(seq, bg="wholeGenome", proteome=proteome)

3.4 Step 3, do test

Test can be done without any change of the symbol pattern or with changes of grouped peptides by such as charge, chemistry, hydrophobicity and etc.

t0 <- testDAU(seq, bg)
t1 <- testDAU(seq, bg, group="classic")
t2 <- testDAU(seq, bg, group="charge")
t3 <- testDAU(seq, bg, group="chemistry")
t4 <- testDAU(seq, bg, group="hydrophobicity")

3.5 Step 4, graphical representation results

We can use heatmap or logo to show the results.

dagHeatmap(t0) ##Plot a heatmap to show the results

dagLogo(t0) ##Plot a logo to show the ungrouped results

##Plot a logo to show the classic grouped results
dagLogo(t1, namehash=nameHash(t1@group), legend=TRUE)

##Plot a logo to show the charge grouped results
dagLogo(t2, namehash=nameHash(t2@group), legend=TRUE)

##Plot a logo to show the chemistry grouped results
dagLogo(t3, namehash=nameHash(t3@group), legend=TRUE)

##Plot a logo to show the hydrophobicity grouped results
dagLogo(t4, namehash=nameHash(t4@group), legend=TRUE)