motifx: Find overrepresented sequence motifs

In omarwagih/rmotifx: An iterative statistical approach to the discovery of biological sequence motifs

Description

Find overrepresented sequence motifs

Usage

motifx(fg.seqs, bg.seqs, central.res = "ST", min.seqs = 20,
  pval.cutoff = 1e-06, verbose = F, perl.impl = F)

Arguments

fg.seqs	Foreground k-mer sequences in a pre-aligned format. All k-mers must have same lengths.
bg.seqs	Background k-mer sequences in a pre-aligned format. All k-mers must have same lengths.
central.res	Central amino acid of the k-mer. Sequences without this amino acid in the centre position are filtered out. This can be one or more letter. For example, 'S', 'ST', 'Y', or 'STY'.
min.seqs	This threshold refers to the minimum number of times you wish each of your extracted motifs to occur in the data set. An occurrence threshold of 20 usually is appropriate, although this parameter may be adjusted to yield more specific or less specific motifs.
pval.cutoff	The p-value threshold for the binomial probability. This is used for the selection of significant residue/position pairs in the motif. A threshold of 0.000001 is suggested to maintain a low false positive rate in standard protein motif analyses.
verbose	If true, motifx will show textual details of the steps while running.
perl.impl	The original implementation of motifx in perl, P-values below 1e-16 cannot be computed and are thus set to zero. Motifx therefore sets any P-values of zero to the minimal P-value of 1e-16. In R, the minimal P-value is much lower (depending on the machine). If this option is set to TRUE, P-values with a value of zero are set to 1e-16, as in perl. Otherwise, the R P-value minimum will be used. For results identical to that of the webserver implementation, set to TRUE.

Value

Data frame with seven columns containing overrepresented motifs. Motifs are listed in the order in which they are extracted by the algorithm, not with regard to statistical significance. Thus it should not be assumed that a motif found at a higher position in the list is more statistically significant than a motif found at a lower position in the list. The columns are as follows:

motif

The overrepresented motif

score

The motif score, which is calculated by taking the sum of the negative log probabilities used to fix each position of the motif. Higher motif scores typically correspond to motifs that are more statistically significant as well as more specific

fg.matches

Frequency of sequences matching this motif in the foreground set

fg.size

Total number of foreground sequences

bg.matches

Frequency of sequences matching this motif in the background set

bg.size

Total number of background sequences

fold.increase

An indicator of the enrichment level of the extracted motifs. Specifically, it is calculated as (foreground matches/foreground size)/(background matches/background size).

Examples

# Get paths to sample files
fg.path = system.file("extdata", "fg-data-ck2.txt", package="rmotifx")
bg.path = system.file("extdata", "bg-data-serine.txt", package="rmotifx")

# Read in sequences
fg.seqs = readLines( fg.path )
bg.seqs = readLines( bg.path )

# Find overrepresented motifs
mot = motifx(fg.seqs, bg.seqs, central.res = 'S', min.seqs = 20, pval.cutoff = 1e-6)

# View results
head(mot)

omarwagih/rmotifx documentation built on May 24, 2019, 1:50 p.m.