calculateSpike: Detect Spike

In strvalidator: Process Control and Validation of Forensic STR Kits

Detect Spike

Description

Detect samples with possible spikes in the DNA profile.

Usage

calculateSpike(
  data,
  threshold = NULL,
  tolerance = 2,
  kit = NULL,
  quick = FALSE,
  debug = FALSE
)

Arguments

data	data.frame with including columns 'Sample.Name', 'Marker', 'Size'.
threshold	numeric number of peaks of similar size in different dye channels to pass as a possible spike (NULL = number of dye channels minus one to allow for one unlabeled peak).
tolerance	numeric tolerance for Size. For the quick and dirty rounding method e.g. 1.5 rounds Size to +/- 0.75 bp. For the slower but more accurate method the value is the maximum allowed difference between peaks in a spike.
kit	string or numeric for the STR-kit used (NULL = auto detect).
quick	logical TRUE for the quick and dirty method. Default is FALSE which use a slower but more accurate method.
debug	logical indicating printing debug information.

Details

Creates a list of possible spikes by searching for peaks aligned vertically (i.e. nearly identical size). There are two methods to search. The default method (quick=FALSE) method that calculates the distance between each peak in a sample, and the quick and dirty method (quick=TRUE) that rounds the size and then group peaks with identical size. The rounding method is faster because it uses the data.table package. The accurate method is slower because it uses nested loops - the first through each sample to calculate the distance between all peaks, and the second loops through the distance matrix to identify which peaks lies within the tolerance. NB! The quick method may not catch all spikes since two peaks can be separated by rounding e.g. 200.5 and 200.6 becomes 200 and 201 respectively.

Value

data.frame

See Also

data.table

strvalidator documentation built on July 26, 2023, 5:45 p.m.