R: create a PWM (Position Weight Model) model given a initial...

refinePWMMotif {motifRG}

R Documentation

create a PWM (Position Weight Model) model given a initial set of motif matches and input sequences

Description

Create a PWM model given a initial set of motif matches and input sequences

Usage

  refinePWMMotif(motifs=NULL, seqs,  pwm.ld= NULL, max.iter=50,
tol=10^-4, mod="oops", null=rep(0.25, 4),pseudo=1, weights=rep(1,
length(seqs)), motif.weights=NULL)

Arguments

`motifs`	The initial set of motif matches. character vector or DNAStringSet object
`seqs`	Input sequences. character vector or DNAStringSet object
`pwm.ld`	The initial PWM matrixes in logodds transformation. Either "motifs" or "pwm.ld" is not NULL
`max.iter`	Maximum number of iterations for refinement
`tol`	Convergence criteria. The percentage of total PWM scores improvement required for convergence.
`mod`	Motif ocurrence model. If `mod=="oops"`, assume one motif match per sequence. If `mod=="zoops"`, assume zero or one motif match per sequence.
`null`	A numeric vector specifying the background model
`pseudo`	Pseudo counts for PWM construction
`weights`	a numeric vector specifying the weights for all sequences. Default: 1 for all sequences
`motif.weights`	a numeric vector specifying the weights for initial sets of motifs. Default: NULL

Value

Return a list with two elements:

model

a list with two elements. "prob": PWM model, sum of columns add to 1. "logodd": PWM model in logodds form, log2 of original matrix substract the background model

`match`	a data.frame specifying the motif matches in each sequence. Columns are: "match": the sequence of the match, "score": PWM score, "strand", the strand of the match in the input sequence, "pos": start position of the motif match. If multiple matches are allowed, then "seq.id" specifies the index of the input sequence for the motif match.
`score`	Total PWM score of the motif matches

Examples

ctcf.seq <- readDNAStringSet(system.file("extdata","ctcf.fa", package="motifRG"))
data(ctcf.motifs)
### refine PWM model based on motif matches
pwm.match <- refinePWMMotif(ctcf.motifs$motifs[[1]]@match$pattern, ctcf.seq)
### plot traditional motif logo
library("seqLogo")
seqLogo(pwm.match$model$prob)
### plot dinucleotide motif logo
plotMotif(pwm.match$match$pattern)
### automatically extend PWM model 
pwm.match.extend <-refinePWMMotifExtend(ctcf.motifs$motifs[[1]]@match$pattern, ctcf.seq)
### plot the new motif matches
plotMotif(pwm.match.extend$match$pattern)