PWMUnscaled {PWMEnrich} | R Documentation |
The PWM function from Biostrings without unit scaling
PWMUnscaled(x, id = "", name = "", type = c("log2probratio", "prob"), prior.params = c(A = 0.25, C = 0.25, G = 0.25, T = 0.25), pseudo.count = prior.params, unit.scale = FALSE, seq.count = NULL)
x |
the integer count matrix representing the motif, rows as nucleotides |
id |
a systematic ID given to this PWM, could include the source, version, etc |
name |
the name of the transcription factor (TF) to which the PWM corresponds to |
type |
the type of PWM calculation, either as log2-odds, or posterior probability (frequency matrix) |
prior.params |
the pseudocounts for each of the nucleotides |
pseudo.count |
the pseudo-count values if different from priors |
unit.scale |
if to unit.scale the pwm (default is no unit scaling) |
seq.count |
if x is a normalised PFM (i.e. with
probabilities instead of sequence counts), then this
sequence count will be used to convert |
By default the Biostrings package scales the log-odds score so it is within 0 and 1. In this function we take a more traditional approach with no unit scaling and offer unit scaling as an additional parameter.
See ?PWM from Biostrings for more information on input arguments.
a new PWM object representing the PWM
if(require("PWMEnrich.Dmelanogaster.background")){ data(MotifDb.Dmel.PFM) ttk = MotifDb.Dmel.PFM[["ttk"]] PWMUnscaled(ttk, id="ttk-JASPAR", name="ttk") # make a PWM with uniform background PWMUnscaled(ttk, id="ttk-JASPAR", name="ttk", prior.params=c("A"=0.2, "C"=0.3, "G"=0.3, "T"=0.2)) # custom background prior = getBackgroundFrequencies("dm3", quick=TRUE) # get background for drosophila (quick mode on a reduced dataset) PWMUnscaled(ttk, id="ttk-JASPAR", name="ttk", prior.params=prior) # convert using genomic background }