R package that does variance component estimation and genetic parameters for linear mixed effect models ( ie animal models). Variance components can be for individual and/or maternal genetic variation, and can be for additive, dominance, epistatic or sexlinked inheritance. Components can be class specific.

Most other packages use iterative likelihood maximization techniques for variance component estimation. dmm uses a dyadic model which in effect reduces variance component estimation to a regression problem. This has the advantages of being non-iterative and of allowing any of the standard regression techniques to be used. The package currently offers least squares, partial least squares and robust regression. The results obtained with dmm are equivalent to MINQUE and bias-corrected-ML estimates, if least squares regression is used.

• From CRAN

• see the package page for the latest release of dmm on CRAN, download the package source, and install with R CMD INSTALL ...

• install the package directly in R with install.packages("dmm")

• From GitHub

• clone or download the latest development version. The version found on GitHub may sometimes be a later development than the CRAN release

• working with quantitative genetics in the R statistical and programming environment
• analysis of small multi-trait datasets with pedigree information
• individual, maternal, and cohort environmental component estimates and standard errors
• individual and maternal additive, dominance, epistatic, and sex-linked genetic component estimates and standard errors
• cross-effect and cross-trait covariance components
• multicollinearities among the components
• genetic parameters (ie proportion of variance and correlation) and standard errors for all fitted components
• genetic response to phenotypic selection for individual additive and maternal additive cases with autosomal and sexlinked components
• data preparation tools
• S3 methods to organize output
• test example datasets
• alternative approach to iterative ML and REML estimation procedures
• component estimates equivalent to MINQUE (after fixed effects by OLS) and bias-corrected-ML (after fixed effects by GLS)
• multi-trait or traitspairwise or traitsblockwise analyses to deal with unequal replication across traits
• class-specific component and parameter estimates
• variance components between maternal or paternal founderlines
• dmm was developed for analysis of sheep breeding data. Workers from other fields would certainly be able to broaden its approach, and contributions would be welcome.

The dmm package relies on the nadiv package for generating relationship matrices. The author of nadiv (Matthew Wolak) has assisted with its use from dmm particularly in relation to sexlinked inheritance and sex-specific genetic parameters.

• Currently addressing the issue of sex-specific and fixed-effect-specific genetic parameters

• Needs attention to the following

• efficiency considerations in construction of dyadic model equations

• a better interface to pls or some other approach to dealing with multicollinearities among the components

• memory usage

The document dmmOverview.pdf gives a complete coverage of all aspects of dmm including worked examples. It is most easily obtained at https://github.com/nevillejackson/dmm/tree/master/dmm/inst/doc and is obtained automatically if you install the CRAN package.