Welcome to DMU

**DMU** is a software packaged for analyzing multivariate linear mixed models.

Development of the DMU software started in the late 1980's in the cattle genetic group at the National
Institute of Animal Science. After several reorganization and merging, this group is now part of
Center for Quantitative Genetics and Genomics (QGG), Faculty of Technical Sciences, Aarhus University.

The main developer are: Per Madsen (per.madsen@qgg.au.dk) and Just Jensen (just.jensen@qgg.au.dk)

Note

DMU is under active development.

• Preface
  • History
  • New setup
  • Conditions for use
• Installation
• DMU executable
  • DMU
• Type of analysis
  • Type 1: AI-REML estimation of co-variance component (previous DMUAI)
  • Type 2: MCMC sampling of dispersion and location parameters (previous RJMC)
  • Type 11: BLUP in core (previous DMU4)
  • Type 12: BLUP-PCG iteration on data (previous DMU5)
• Driverfile key words
  • $COMMENT
  • $ANALYSES
  • $DATA
  • $VARIABLE
  • $MODEL
  • $GLMM
  • $GLMM_PRED
  • $REDUCE
  • $VAR_STR
  • $VAR_REST
  • $MIXTURE
  • $PRIOR
  • $PRECOND
  • $SOLUTION
  • $SOL_COV
  • $SOL_PEV
  • $RESIDUAL
  • $BLUP PCG
  • $AI-REML
  • $MCMC
• Datafile
• Model
  • TRAITS
  • ABSORB
  • CLASS Effects
  • RANDOM
  • REGRES
  • NOCOV
• Weighted
  • Weighted
• Variance structure
  • PED
  • GREL
  • DOM
  • COR
  • GROUP
  • ABS_QTL
• (Co)-variances
  • Starting values, priors, “true” values
• Mandatory and optional input by analyze type
  • Optional input - analyze type 1 (AI-REML)
  • Mandatory and optional input - analyze type 2 (MCMC)
  • Optional input - analyze type 12 (BLUP PCG iteration on data)