Phasing, Pedigree Reconstruction, Sire Imputation and Recombination Events Identification for Half-sib Families

Description

Identification of recombination events, haplotype reconstruction and sire imputation using half-sib family SNP data.

Details

Main Functions:
bmh: Block partitioning
ssp: Sire inference
aio: Phasing
imageplot: Image plot of the block structure
rpoh: Reconstruct pedigree based on opposing homozygote

Auxiliary Functions
hss: Half-sib family splitter
cs: Chromosome splitter
para: Parallel data analysis

Note: These functions can be used to analyse large datasets.

Author(s)

Mohammad H. Ferdosi <mferdosi@une.udu.au>, Cedric Gondro <cgondro2@une.edu.au> Maintainer: Mohammad H. Ferdosi <mferdosi@myune.udu.au>

References

Ferdosi, M. H., Kinghorn, B. P., van der Werf, J. H., & Gondro, C (2013). Effect of genotype and pedigree error on detection of recombination events, sire imputation and haplotype inference using the hsphase algorithm. In Proc. Assoc. Advmt. Anim. Breed. Genet (Vol. 20, pp. 546-549). AAABG; Napier, New Zealand.

Ferdosi, M. H., Kinghorn, B. P., van der Werf, J. H., & Gondro, C. (2014). Detection of recombination events, haplotype reconstruction and imputation of sires using half-sib SNP genotypes. Genetics, selection, evolution: GSE, 46(1), 11.

Ferdosi, M. H., Kinghorn, B. P., van der Werf, J. H., Lee, S. H., & Gondro, C. (2014). hsphase: an R package for pedigree reconstruction, detection of recombination events, phasing and imputation of half-sib family groups. BMC Bioinformatics, 15(1), 172.

Ferdosi, M. H., & Boerner, V. (2014). A fast method for evaluating opposing homozygosity in large SNP data sets. Livestock Science.

Examples

genotype <- matrix(c(
  0,0,0,0,1,2,2,2,0,0,2,0,0,0,
  2,2,2,2,1,0,0,0,2,2,2,2,2,2,
  2,2,2,2,1,2,2,2,0,0,2,2,2,2,
  2,2,2,2,0,0,0,0,2,2,2,2,2,2,
  0,0,0,0,0,2,2,2,2,2,2,0,0,0), ncol = 14, byrow = TRUE) 
ssp(bmh(genotype), genotype)
aio(genotype)
imageplot(bmh(genotype), title = "ImagePlot example")
rplot(genotype, c(1:14))

Calculate Genotypic Distances

Description

Calculates a symmetric matrix of distances between genotypes, based on a given genotype matrix. Each row in the 'GenotypeMatrix' represents a genotype, and each column represents a marker. The genotype is coded as 0 for AA, 1 for AB, and 2 for BB. Use 9 to represent missing data.

Usage

.fastdist(GenotypeMatrix)

Arguments

Value

Returns a symmetric matrix of distances between the genotypes specified in the 'GenotypeMatrix'. Row and column names of the returned matrix correspond to the row names of the 'GenotypeMatrix'.

Examples

# Simulate genotype data for 40 individuals across 1000 SNPs
# genotypes <- simulateHalfsib(numInd = 40, numSNP = 1000, recbound = 0:6, type = "genotype")
# Calculate the distance matrix
# dist_matrix <- fastdist(genotypes)
# Display the distance matrix
# print(dist_matrix)

Calculate Minor Allele Frequency (MAF)

Description

This function calculates the minor allele frequency (MAF) for a given single nucleotide polymorphism (SNP) data. The SNP data should be coded numerically: 0 for homozygous for the first allele (AA), 1 for heterozygous (AB), and 2 for homozygous for the second allele (BB). Missing data should be coded as 9.

Usage

.maf(snp)

Arguments

Value

A numeric value representing the minor allele frequency (MAF) for the SNP data provided.

Examples

snp_data <- c(0, 0, 1, 2, 2, 9)
maf_value <- .maf(snp_data)
print(maf_value)

Simulate Half-Sibling Genotypes

Description

This function simulates genotypes for a set of half-siblings based on specified parameters, including the number of individuals, the number of SNPs, recombination boundaries, and the type of data to return. It generates a sire genotype, maternal half-sib genotypes, and combines these to simulate offspring genotypes, optionally returning phased genotypes based on recombination events.

Usage

.simulateHalfsib(
  numInd = 40,
  numSNP = 10000,
  recbound = 0:6,
  type = "genotype"
)

Arguments

Value

Depending on the type parameter, this function returns a matrix of simulated genotypic data for half-siblings. If type is "genotype", it returns unphased genotypic data; otherwise, it returns phased genotypic data.

Examples

sim_genotypes <- .simulateHalfsib(numInd = 40, numSNP = 10000, recbound = 0:6, type = "genotype")
dim(sim_genotypes) # Should return 40 rows (individuals) and 100 columns (SNPs)

All-in-one Phasing

Description

Phasing of a half-sib family group.

Usage

aio(genotypeMatrix, bmh_forwardVectorSize = 30, bmh_excludeFP = TRUE,
bmh_nsap = 3, output = "phase")

Arguments

Details

This function calls the bmh, ssp and phf functions.

Value

Returns a list of matrices. The first element (phasedHalfsibs) is a matrix with two rows (phased haplotypes) per individual (first paternal and second maternal). Data in format 0 (A), 1 (B) and 9 (unphased or missing). The second (sireHaplotype) and third (blockStructure) elements are the same as the output of ssp and bmh.

Note

Only this function needs to be called to phase a half-sib family. The genotype's matrix must contain individuals from only one half-sib family and one ordered chromosome.

See Also

bmh, ssp and phf

Examples

genotype <- matrix(c(       # Define a Half-sib Genotype Matrix
  2,1,0,                    # Individual 1
  2,0,0,                    # Individual 2
  0,0,2                     # Individual 3
  ), byrow = TRUE, ncol = 3)    # There are 3 individulas with three SNPs
  
aio(genotype)               # The genotypes must include only one half-sib family and one chromosome

Block Partitioning

Description

Identifies the block structure (chromosome segments) in the half-sib family that each individual inherited from its sire.

Usage

bmh(GenotypeMatrix, forwardVectorSize = 30, excludeFP = TRUE, nsap = 3)

Arguments

Value

Returns a matrix of the blocking structure that contains 1s, 2s and 0s. 1s and 2s are the two sire strands. The choice of strand is arbitrary for each chromosome and not consistent across chromosomes. 0s indicate regions of unknown origin.

Note

The genotype's matrix must contain individuals from only one half-sib family and one ordered chromosome.

See Also

ssp, phf, aio and imageplot

Examples

genotype <- matrix(c(
0,2,1,1,1,
2,0,1,2,2,
2,2,1,0,2,
2,2,1,1,1,
0,0,2,1,0), ncol = 5, byrow = TRUE) 

(result <- bmh(genotype))

Crossover Detection

Description

Detect all possible crossover events.

Usage

co(genotypeMatrix)

Arguments

Value

Returns a matrix with the number of crossover events for each site.

Examples

genotype <- matrix(c(           # Define a Half-sib Genotype Matrix
  2,1,0,                        # Individual 1
  2,0,2,                        # Individual 2
  0,0,2                         # Individual 3
  ), byrow = TRUE, ncol = 3)    # There are 3 individuals with three SNPs
  
co(genotype)               

Chromosome Splitter

Description

This function splits the genotypes list generated by hss into the different chromosomes based on a map file and orders SNP based on chromosomal position.

Usage

cs(halfsib, mapPath, separator = " ")

Arguments

Details

The map file should include only the chromosomes that will be analyzed. For example, the Y and X chromosomes should be excluded (and others optionally). Names of each element in the list can be used for further categorization. The header must be "Name Chr Position".

Value

Returns a list of matrices, the number of elements in this list is the number of half-sib families multiplied by the number of chromosomes.

Examples

# Please run demo(hsphase)

Example of Genotype Data Set

Description

This data set serves as an example of a genotype matrix intended for use with the hsphase package.

Usage

data(genotypes)

Format

The data set is a genotype matrix with specific structure, including:

• Columns: Represent Single Nucleotide Polymorphisms (SNPs). Each column corresponds to a specific SNP.

• Rows: Represent individual animals. Each row corresponds to the genotypic data for a single animal across various SNPs.

Haplotype Blocks of Phased Data

Description

Creates a blocking structure matrix of the half-sib family based on phased data of the sire and half-sib family.

Usage

hbp(PhasedGenotypeMatrix, PhasedSireGenotype, strand = "auto")

Arguments

Value

Returns a matrix where 3 or 4 stands for the SNP originating in, respectively, strands 1 and 2. 0 indicates that the source strand for the SNP is unknown.

Note

The input matrices must only contain individuals from one half-sib family and one ordered chromosome. The strand option should be set to "auto" (default value).

See Also

aio, ssp

Examples

sire <- matrix(c(
  0,0,0,0,0,1,                  # Haplotype one of the sire
  0,1,1,1,1,0                   # Haplotype two of the sire
  ), byrow = TRUE, ncol = 6)
  
haplotypeHalfsib <- matrix(c(
  1,0,1,1,1,1,                  # Individual one, haplotype one
  0,1,0,0,0,0,                  # Individual one, haplotype two
  0,1,1,0,1,1,                  # Individual two, haplotype one
  1,0,0,1,0,0                   # Individual two, haplotype two
  ), byrow = TRUE, ncol = 6)    # 0s and 1s are alelle a and b
  
 hbp(haplotypeHalfsib, sire)

Heatmap of Half-sibs

Description

The hh function creates a heatmap of the half-sib families using the matrix of opposing homozygotes.

Usage

	hh(oh, inferredPedigree, realPedigree, pedOnly = TRUE)

Arguments

Value

Returns the heatmap of the matrix of opposing homozygotes with sidebars colour coded by sires from the inferred and original pedigree.

Author(s)

The fuction uses the colour generated by getcol function in the made4 package (Aedin Culhane).

See Also

ohg and rpoh

Examples

c1h1 <- .simulateHalfsib(numInd = 62, numSNP = 5000)
c1h2 <- .simulateHalfsib(numInd = 38, numSNP = 5000)
Genotype <- rbind(c1h1, c1h2)
oh <- ohg(Genotype) # creating the Opposing Homozygote matrix
hh(oh)

Half-sib Family Splitter

Description

Splits the dataset into half-sib family groups based on a pedigree.

Usage

hss(pedigree, genotype, check = TRUE)

Arguments

Details

Only half-sib groups that have more than 3 individuals will be returned.

Value

Returns a list of numeric matrices, each matrix is a half-sib family.

Note

Pedigree must have at least two columns with sample ids (Column 1) and sire ids (Column 2).

Examples

# Please run demo(hsphase)

Image Plot of Blocking Structure

Description

Create an imageplot of the blocking structure.

Usage

imageplot(x, title, rv = FALSE, ...)

Arguments

Details

White indicates regions of unknown origin, red and blue correspond to the two sire strands.

Author(s)

This is a modified version of a function written by Chris Seidel.
http://www.phaget4.org/R/image_matrix.html

See Also

bmh and aio

Examples

genotype <- matrix(c(
0,2,1,1,1,
2,0,1,2,2,
2,2,1,0,2,
2,2,1,1,1,
0,0,2,1,0), ncol = 5, byrow = TRUE) # each row contains the SNP of individuals
imageplot(bmh(genotype))

Impute of Low Density SNP Marker to High Density (Paternal Strand)

Description

Impute the paternal strand from low density to high density utilising high density sire haplotype.

Usage

impute(halfsib_genotype_ld, sire_hd, bmh_forwardVectorSize = 30, 
bmh_excludeFP = TRUE, bmh_nsap = 3)

Arguments

Value

Return an imputed half-sib matrix.

See Also

bmh, ssp and phf

Example Map File for Genetic Data

Description

This data set is an example of a map file used within the hsphase package to demonstrate the mapping of SNPs to their respective locations on chromosomes.

Usage

data(map)

Format

The data set is formatted as a data frame with the following columns, providing essential information about each SNP:

• Name: The unique identifier or name of the SNP.

• Chr: The chromosome on which the SNP is located.

• Position: The position of the SNP on the chromosome, expressed in base pairs.

Opposing Homozygote Detection

Description

Counts the number of opposing homozygotes for each animal that caused a heterozygus site in the sire.

Usage

ohd(genotypeMatrix, unique_check = FALSE, SNPs = 6000)

Arguments

Value

Returns a vector with the number of heterozygous sites that each sample caused.

Note

This function can be used to identify pedigree errors; i.e., the outliers.

Author(s)

This method is suggested by Bruce Tier <btier@une.edu.au> to identify pedigree errors.

Examples

genotype <- matrix(c(        
  2,1,0,                  
  2,0,0,                  
  0,0,2                   
  ), byrow = TRUE, ncol = 3)    
  
ohd(genotype)             

Matrix of Opposing Homozygotes

Description

Creates a matrix of opposing homozygotes from the genotype matrix.

Usage

ohg(genotypeMatrix)

Arguments

Value

Returns a square matrix (sample X sample) with the pairwise counts of opposing homozygotes.

Note

This function can be slow with a large data set. The fast version of this function will be available after publish of the related manuscript.

Author(s)

Ferdosi, M. H., & Boerner, V. (2014). A fast method for evaluating opposing homozygosity in large SNP data sets. Livestock Science.

See Also

rpoh

Examples

genotype <- matrix(c(        
  2,1,0,                  
  2,0,0,                  
  0,0,2                   
  ), byrow = TRUE, ncol = 3)    
  
ohg(genotype)             

Opposing Homozygotes Plot

Description

Plot the sorted vectorized matrix of Opposing Homozygotes.

Usage

ohplot(oh, genotype, pedigree, check = FALSE)

Arguments

Details

The cut off line shows the edge of most different groups.

See Also

ohg and rpoh

Examples

set.seed(100)
chr <- list()
sire <- list()
set.seed(1)
chr <- list()
for(i in 1:5)
{
	chr[[i]] <- .simulateHalfsib(numInd = 20, numSNP = 5000, recbound = 1:10)
	sire[[i]] <- ssp(bmh(chr[[i]]), chr[[i]])
	sire[[i]] <- sire[[i]][1,] + sire[[i]][2,]
	sire[[i]][sire[[i]] == 18] <- 9
}

Genotype <- do.call(rbind, chr)
rownames(Genotype) <- 6:(nrow(Genotype) + 5)
sire <- do.call(rbind, sire)
rownames(sire) <- 1:5
Genotype <- rbind(sire, Genotype)
oh <- ohg(Genotype)  # creating the Opposing Homozygote matrix
pedigree <- as.matrix(data.frame(c(1:5, 6:(nrow(Genotype))), 
rep = c(rep(0,5), rep(1:5, rep(20,5)))))
ohplot(oh, Genotype, pedigree, check = TRUE)

Parallel Analysis of Data

Description

This function uses the list of matrices (the output of cs) and runs one of the options, on each element of the list, in parallel.

Usage

para(halfsibs, cpus = 1, option = "bmh", type = "SOCK", bmh_forwardVectorSize = 30,
bmh_excludeFP = TRUE, bmh_nsap = 3, pmMethod = "constant")

Arguments

Details

Type of analysis can be bmh, ssp, aio, pm, or rec (refer to pm, rplot and vignette for more information about rec).

Value

Returns a list of matrices with the results (formats specific to the option selected).

Examples

# Please run demo(hsphase)

Example Pedigree Data Set

Description

This dataset provides an example of a pedigree, specifically designed for use with the hsphase package.

Usage

data(pedigree)

Format

The dataset is structured as a data frame with detailed familial relationships, including:

• First Column: Identifiers for half-sibs.

• Second Column: Identifiers for sires.

Fix Pedigree Errors

Description

Tries to link the inferred pedigree from rpoh with the sire IDs in the original pedigree and fix pedigree errors.

Usage

pedigreeNaming(inferredPedigree, realPedigree)

Arguments

Details

This function calls the bmh and recombinations functions to count the number of recombinations in each half-sib group.

Value

Returns the inferred pedigree with the best fit to the sire names used in the original pedigree file.

See Also

rpoh and ohg

Examples

# Please run demo(hsphase)

Half-Sib Family Phasing

Description

Phases the half-sib family by using the blocking structure and imputed sire matrices.

Usage

phf(GenotypeMatrix, blockMatrix, sirePhasedMatrix)

Arguments

Value

Returns a matrix that contains the phased parental haplotypes of the half-sibs. It uses 1, 0 and 9 for A, B and missing.

Note

The genotype matrix must only contain individuals from one half-sib family and one ordered chromosome. This function is used by the aio function for complete phasing of a half-sib group.

See Also

aio

Examples

genotype <- matrix(c(
  2,1,0,
  2,0,0,
  0,0,2), byrow = TRUE, ncol = 3)
block <- bmh(genotype)
phf(genotype, block, ssp(block, genotype))

Probability Matrix

Description

Creates a recombination matrix based on the blocking structure.

Usage

pm(blockMatrix, method = "constant") 

Arguments

Details

This function finds the recombination between two consecutive sites, and marks the recombination site with a 1; if there are unknown sites between two blocks it will also mark these sites with a 1 (constant method) or 1 divided by number of unknown site (relative method).

Examples

genotype <- matrix(c(
0,2,0,1,0,
2,0,1,2,2,
2,2,1,0,2,
2,2,1,1,1,
0,0,2,1,0), ncol = 5, byrow = TRUE) # each row contains the SNP of individuals

(result <- bmh(genotype))
pm(result)

Parent Offspring Group Constructor

Description

Assign offsprings to the parents.

Usage

pogc(oh, genotypeError)

Arguments

Value

Return a data frame with two columns. The first column is the animal ID and the second column is the parent ID.

See Also

ohg, hss and rpoh

Examples

set.seed(100)
chr <- list()
sire <- list()
set.seed(1)
chr <- list()
for(i in 1:5)
{
	chr[[i]] <- .simulateHalfsib(numInd = 20, numSNP = 5000, recbound = 1:10)
	sire[[i]] <- ssp(bmh(chr[[i]]), chr[[i]])
	sire[[i]] <- sire[[i]][1,] + sire[[i]][2,]
	sire[[i]][sire[[i]] == 18] <- 9
}

Genotype <- do.call(rbind, chr)
rownames(Genotype) <- 6:(nrow(Genotype) + 5)
sire <- do.call(rbind, sire)
rownames(sire) <- 1:5
Genotype <- rbind(sire, Genotype)
oh <- ohg(Genotype)  # creating the Opposing Homozygote matrix
pogc(oh, 5)

Read and Check the Genotype File

Description

This function reads and checks genotype files.

Usage

readGenotype(genotypePath, separatorGenotype = " ", check = TRUE)

Arguments

Value

Returns the genotype matrix.

Note

Please refer to vignette for more information.

Examples

# A comprehensive demo and example dataset is available from
# http://www-personal.une.edu.au/~cgondro2/hsphase.html

Recombination Number

Description

Counts the number of recombinations for each individual.

Usage

recombinations(blockMatrix)

Arguments

Value

Returns a vector of recombinations. The number of elements in this vector is equal to the number of individuals, i.e. each element holds the number of recombinations identified for each sample.

See Also

bmh

Examples

genotype <- matrix(c(          
  2,1,0,0,                    
  2,0,2,2,                    
  0,0,2,2,
  0,2,0,0                     
  ), byrow = TRUE, ncol = 4)      
  
recombinations(bmh(genotype))             

Recombination Plot

Description

This function creates a plot which shows the sum of all recombination events across a half-sib family.

Usage

rplot(x, distance, start = 1, end = ncol(x), maximum = 100,
overwrite = FALSE, method = "constant")

Arguments

Examples

genotype <- matrix(c(
0,2,0,1,0,
2,0,1,2,2,
2,2,1,0,2,
2,2,1,1,1,
0,0,2,1,0), ncol = 5, byrow = TRUE) # each row contains the SNP of individuals

rplot(genotype, c(1,2,3,4,8)) 

Reconstruct Pedigree Based on Matrix of Opposing Homozygotes

Description

Reconstructs a half-sib pedigree based on a matrix of opposing homozygotes.

Usage

rpoh(genotypeMatrix, oh, forwardVectorSize = 30, excludeFP = TRUE, nsap = 3,
maxRec = 15, intercept = 26.3415, coefficient = 77.3171, snpnooh, method, maxsnpnooh)

Arguments

Details

Four methods simple, recombinations, calus and manual can be utilized to reconstruct the pedigree.

The following examples show the arguments require for each method.

pedigree1 <- rpoh(oh = oh, snpnooh = 732, method = "simple")
pedigree2 <- rpoh(genotypeMatrix = genotypeChr1, oh = ohg(genotype), maxRec = 10 , method = "recombinations")
pedigree3 <- rpoh(genotypeMatrix = genotype, oh = oh, method = "calus")
pedigree4 <- rpoh(oh = oh, maxsnpnooh = 31662, method = "manual")

Value

Returns a data frame with two columns, the first column is animals' ID and the second column is sire identifiers (randomly generated).

Note

Method can be recombinations, simple, calus or manual. Please refer to vignette for more information.

The sire genotype should be removed before using this function utilizing pogc function.

See Also

bmh and recombinations

Examples

# Please run demo(hsphase)

Sire Imputation and Phasing

Description

Infer (impute) and phase sire's genotype based on the block structure matrix (recombination blocks) and homozygous sites of the half-sib genotype matrix.

Usage

ssp(blockMatrix, genotypeMatrix)

Arguments

Value

Returns a matrix (Imputed Sire) with two rows one for each haplotype of the sire (columns are SNP in the order of the genotype matrix). Alleles are coded as 0 (A) and 1 (B). Alleles that could not be imputed are coded as 9.

See Also

phf, aio and imageplot

Examples

genotype <- matrix(c(
0,2,1,1,1,
2,0,1,2,2,
2,2,1,0,2,
2,2,1,1,1,
0,0,2,1,0), ncol = 5, byrow = TRUE) # each row contains the SNP of individuals

(result <- ssp(bmh(genotype), genotype))