getExpInterval: Identify the interval of relative substitution frequencies...

In wavClusteR: Sensitive and highly resolved identification of RNA-protein interaction sites in PAR-CLIP data

Description

Identifies the interval/support of relative substitution frequencies (RSFs) dominated by the second model component, i.e. by the probability of being induced by the experimental procedure. In addition, this function can be used to generate diagnostic plots of the model fit, representing (i) model densities and log odds ratio (ii) the posterior class probability, i.e. the probability of a given observation being generated by experimental induction.

Usage

getExpInterval(model, bayes = TRUE, leftProb, rightProb, plot = TRUE)

Arguments

model	A list containing the model as returned by the function fitMixtureModel
bayes	Logical, if TRUE the Bayes classifier (cutoff at posterior class probabilities >= 0.5) is applied. If FALSE, custom cutoff values should be provided through leftProb and rightProb. Default is TRUE.
leftProb	Numeric, the posterior probability corresponding to the left boundary (start) of the high confidence RSF interval.
rightProb	Numeric, the posterior probability corresponding to the right boundary (end) of the high confidence RSF interval.
plot	Logical, if TRUE diagnostics plot showing the model components, log odds and the computed posterior with highlighted identified RSF interval are returned.

Value

A list with two numeric slots, corresponding to the extremes of the RSF interval (RSF support).

supportStart	start of the high confidence RSF interval
supportEnd	end of the high confidence RSF interval

Author(s)

Federico Comoglio and Cem Sievers

References

Sievers C, Schlumpf T, Sawarkar R, Comoglio F and Paro R. (2012) Mixture models and wavelet transforms reveal high confidence RNA-protein interaction sites in MOV10 PAR-CLIP data, Nucleic Acids Res. 40(20):e160. doi: 10.1093/nar/gks697

Comoglio F, Sievers C and Paro R (2015) Sensitive and highly resolved identification of RNA-protein interaction sites in PAR-CLIP data, BMC Bioinformatics 16, 32.

See Also

fitMixtureModel getHighConfSub estimateFDR

Examples

data( model )

#default
support <- getExpInterval( model = model, bayes = TRUE, plot = TRUE )
support

#custom interval (based, e.g. on visual inspection of posterior class probability 
# or evaluation of FDR using the estimateFDRF function)
support <- getExpInterval( model = model, leftProb = 0.2, rightProb = 0.7, plot = TRUE )
support

wavClusteR documentation built on Nov. 8, 2020, 6:54 p.m.