Jacquard (version 1.0.2)

Jacquard.cls: Estimation of Jacquard coefficients by Constrained Least Squares

Description

Function Jacquard.cls estimates the nine condensed Jacquard coefficients of a pair of individuals using their joint genotype probabilities and the allele frequencies using constrained least squares.

Usage

Jacquard.cls(Xlist, mafvec = NULL, eps = 1e-06, delta.init = runif(9), Mavg = NULL,
             inner.iter = 1000, outer.iter = 1000, verbose = TRUE)

Value

A list object with fields:

delta

A list with nine matrices of estimates of pairwise Jacquard coefficients.

convergence

A matrix with the convergence status for each pair (0 = converged; 1 = not converged).

Arguments

Xlist

A list object with nine fields containing the matrices with joint genotype counts.

mafvec

A vector with the minor allele frequencies for all genetic variants.

eps

Tolerance criterion for the solver (eps=1e-06 by default).

delta.init

Initial vector of estimates for the nine condensed Jacquard coefficients.

Mavg

A nine by nine matrix of conditional probabilities, allele frequency dependent. This matrix is calculated by Jacquard.cls if it is not specified.

inner.iter

Maximum number of inner iterations for the solver (1000 by default).

outer.iter

Maximum number of outer iterations for the solver (1000 by default).

verbose

Print output on the progress of the algorithm if verbose=TRUE.

Author

Jan Graffelman ([e-mail])

Details

Function Jacquard.cls relies on the solver solnp from the Rsolnp package.

References

Graffelman, J., Weir, B.S. and Goudet, J. (2024) Estimation of Jacquard's genetic identity coefficients with bi-allelic variants by constrained least-squares. Preprint at bioRxiv tools:::Rd_expr_doi("10.1101/2024.03.25.586682").

Ghalanos, A. and Theussl, S. (2015) Rsolnp: General Non-linear Optimization Using Augmented Lagrange Multiplier Method. R package version 1.16. https://cran.r-project.org/package=Rsolnp

See Also

JointGenotypeCounts

Examples

data(SimulatedPedigree)
Xgen <- as.matrix(SimulatedPedigree[,6:ncol(SimulatedPedigree)])

data(GTC)

mafvec <- mafvector(Xgen)

ii <- 1:3

GTCsubset <- list(length = 9)
for (k in 1:9) {
  GTCsubset[[k]] <- matrix(numeric(3^2), ncol = 3)
  GTCsubset[[k]] <- GTC[[k]][ii,ii]
}

output <- Jacquard.cls(GTCsubset,mafvec=mafvec,
                       eps=1e-06)
Delta.cls <- output$delta
print(Delta.cls)
print(output$convergence)

# A particular estimate of a Jacquard coefficient for a particular pair can
# be extracted from Delta.cls
#
# E.g., Delta\_9 of the first pair of individuals (1,2) can be extracted by
#

D9_12 <- Delta.cls[[9]][1,2]