peakpick: Identifies smooth peaks in series of numbers.
Description
This algorithm detects peaks, smooth bumps in series of numbers. This
algorithm should not be used for series containing brief spikes. Consider
filtering/smoothing your data before using this algorithm. Please refer to
the paper by Weber et al. for more details.
Usage
peakpick(mat, neighlim, deriv.lim = 0.04, peak.min.sd = 0.5, peak.npos = 10L, mc.cores = 1)
Arguments
mat
matrix of series with series organized columnwise
neighlim
integer limit for how far apart peaks must be. Peak pairs
closer than or equal to neighlim to each other have the lesser peak
eliminated.
deriv.lim
numeric upper limit for the estimatied derivative for a
point to be considered for a peak call
peak.min.sd
numeric minimum number of standard deviations for a peak
to rise above the mean of its immediate vicinity in order to be considered
for a peak call
peak.npos
integer peak standard deviations and means will be estimated
plus/minus npos positions from peak
mc.cores
the number of cores to perform this computation
Value
boolean matrix with dimensions of mat representing peaks
References
Weber, C.M., Ramachandran, S., and Henikoff, S. (2014).
Nucleosomes are context-specific, H2A.Z-modulated barriers to RNA
polymerase. Molecular Cell 53, 819-830.