# predict_IBD: Predict IBD In Alethere/SmoothDescent: Application of map-based genotype correction algorithm Smooth Descent

## Description

IBD prediction algorithm, performs a distance-informed weighted average using a genetic map. Weights are calculated as a proportion of the non-recombination probability between two markers, which are estimated using mapping functions (Kosambi's, Haldane's or Morgan's). Low-informative IBD probabilities are ignored, by default probabilities between 0.3 and 0.7 are considered low-informative.

## Usage

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38``` ```predict_IBD( IBD, map, interval = 10, method = "kosambi", non_inf = c(0.3, 0.7), pred_points = NULL ) ## S3 method for class 'list' predict_IBD( IBD, map, interval = 10, method = "kosambi", non_inf = c(0.3, 0.7), pred_points = NULL ) ## S3 method for class 'matrix' predict_IBD( IBD, map, interval = 10, method = "kosambi", non_inf = c(0.3, 0.7), pred_points = NULL ) ## S3 method for class 'numeric' predict_IBD( IBD, map, interval = 10, method = "kosambi", non_inf = c(0.3, 0.7), pred_points = NULL ) ```

## Arguments

 `IBD` observed IBD. Can be given as a vector containing the observed IBDs of one individual, a single matrix (markers x individual) or a list of matrices. Matrices must contain rownames and columnames. `map` data.frame containing "marker" and "position" columns at least. `interval` numeric indicating the interval size to be used in prediction. `method` string indicating the mapping function used to calculate weights. Either "kosambi", "haldane" or "morgan". Kosambi takes into account chiasma interference, Haldane is the most common mapping function and Morgan assumes a linear relationship between distance and recombination (is inaccurate for large distances). `non_inf` numeric vector of two digits, containing the lower and upper bound for probabilities to be considered non-informative. These probabilities will be ignored during prediction. By default 0.3 and 0.7. Symmetric and stringent thresholds are recommended. `pred_points` numeric, number of points to use for IBD prediction. If NULL, all points in map\$position are used, otherwise n equally spaced points are used. Greatly improves efficiency if the number of markers is very large.

## Value

the output will have the same format as the input given. That is, an input IBD vector will return an IBD vector, a matrix will return a matrix and a list of matrices will return a list of matrices.

## Methods (by class)

• `list`: Function for a list of matrices

• `matrix`: Function for a single matrix

• `numeric`: Function for a numeric vector

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12``` ```data("genotype") geno <- geno[,-1:-2] #we take out parental genotypes data("homologue") data(map) IBD <- calc_IBD(geno,hom[,1:2],hom[,3:4], ploidy = 2) #One homologue of one individual pred <- predict_IBD(IBD[[1]][,1], map) #One homologue for all individuals pred <- predict_IBD(IBD[[1]], map) #Or all homologues pred <- predict_IBD(IBD, map) ```

Alethere/SmoothDescent documentation built on Jan. 14, 2022, 9:40 p.m.