get_smuce: Segmentation Algorithm to Estimate Breakpoints in the...

Description Usage Arguments Value Author(s) References See Also Examples

View source: R/get_smuce.R

Description

First, the recombination rates per segment are computed based on the regression model (generalized additive models) as well as the bias correction. Consequently, we apply SMUCE (simultaneous multiscale change-point estimator) of Frick (2014) and Futschik et al. (2014) to estimate locations and breakpoints in the recombination map. Under a specific type-I error probability alpha the number of distinct segments with respect to the recombination rate is not overestimated.

Usage

1
get_smuce(help, segs, alpha, ll, quant = 0.35, rescale, constant, demography, regMod)

Arguments

help

a matrix containing a set of summary statistics is calculated in the function summary_statistics. These values are used in the regression model to calculate the (constant) recombination rates.

segs

A (non-negative) integer which reflects the number of segments considered. It is calculated in the program based on the user-defined segLength.

alpha

A value from the interval (0,1) for the type-I error probability used in the segmentation algorithm. We recommend to use 0.05. We enabled to estimate the recombination map efficiently (without recalculating all summary statistics) under several type-I errors when LDJump is applied with a vector of type-I error probabilities.

ll

A (non-negative) integer which reflects the total sequence length of the sequences under study.

quant

A value between 0.1 and 0.5 with 0.05 distances in between which reflects the quantile used in the quantile regression. We recommend to use the 0.35 quantile.

rescale

an optional logical value: if TRUE, it rescales the sequence length of the output of SMUCE to a range from 0 to 1.

constant

an optional logical value: by default FALSE estimating variable recombination rates. If TRUE, the constant recombination rate for the full sequence is estimated.

demography

an optional character value: by default an empty string ("") indicates that the recombination rate estimation is estimated under neutrality. If "b" the regression model estimated based on samples from populations under a bottleneck is used. If "g" the regression model estimated based on samples from populations under population growth is used. If "d", the regression model estimated based on samples from populations under demography (combination of samples of under growth and bottleneck) is used.

regMod

an optional character string: for the default empty string "" *LDJump* uses an existing regression model (constant population size or simple demography example, depending on demography). In oder to use the regression model estimated by user input demography, then this variable should equal to the name of the regression object. Please see the examples for more details.

Value

seq.full.cor

The final estimate of the recombination map. Depiction with plot-function of stepR package.

pr.full.cor

A vector of (constant) estimates of the recombination rate per segment.

Author(s)

Philipp Hermann philipp.hermann@jku.at, Andreas Futschik

References

Frick, K., Munk, A., and Sieling, H. (2014). Multiscale change-point inference. Journal of the Royal Statistical Society: Series B, 76(3), 495–580.

Futschik, A., Hotz, T., Munk, A., and Sieling, H. (2014). Multiscale DNA partitioning: Statistical evidence for segments. Bioinformatics, 30(16), 2255–2262.

Hermann, P., Heissl, A., Tiemann-Boege, I., and Futschik, A. (2019), LDJump: Estimating Variable Recombination Rates from Population Genetic Data. Mol Ecol Resour. doi:10.1111/1755-0998.12994.

See Also

LDJump, vcfR_to_fasta, getPhi, summary_statistics, stepFit, rq, gam

Examples

1
2
3
##### Do not run these examples                           #####
##### In LDJump.R the function is called as follows       #####
##### get_smuce(help, segs, alpha,ll,list.quantile.regs)  #####

PhHermann/LDJump documentation built on Nov. 16, 2019, 12:53 p.m.