find_valleyBySpline: Find major valleys after spline smoothing

In jianhong/InPAS: Identify Novel Alternative PolyAdenylation Sites (PAS) from RNA-seq data

Find major valleys after spline smoothing

Description

Find major valleys after spline smoothing

Usage

find_valleyBySpline(
  x,
  ss,
  se = length(x),
  nknots = ceiling((se - ss + 1)/1000 * 10),
  n = -1,
  min.dist = 200,
  filter.last = TRUE,
  DIST2END = 1200,
  plot = FALSE
)

Arguments

x	A vector of numeric(n), containing MSEs for a given range
ss	An positive integer, search start site relative to the leftmost base
se	An positive integer, search end site relative to the leftmost base
nknots	An positive integer, the number of knots for smoothing using splinestats::smooth.spline(). By default, set to 10 knots per kb.
n	An integer, specifying the number of location where MSE are local minima (candidate CP sites). If set to -1, return all candidate CP sites.
min.dist	An integer, minimal distance allowed between two adjacent candidate CP sites otherwise collapsed by selecting the one with lower MSE.
filter.last	A logical(1), whether to filter out the last valley, which is likely the 3' end of the longer 3' UTR if no novel distal CP site is detected and the 3' end excluded by setting cutEnd/search_point_END is small.
DIST2END	An integer, specifying a cutoff of the distance between last valley and the end of the 3' UTR (where MSE of the last base is calculated). If the last valley is closer to the end than the specified distance, it will be not be considered because it is very likely due to RNA coverage decay at the end of mRNA. Default is 1200. User can consider a value between 1000 and 1500, depending on the library preparation procedures: RNA fragmentation and size selection.
plot	A logical(1), whether to plot the MSE profile and the candidate valleys.

Value

A vector of integer.

jianhong/InPAS documentation built on Nov. 5, 2024, 7:45 a.m.