Decoding mixed model equations using LMMonBoard

High-throughput Phenotyping Driven Quantitative Genetics @CMA-FCT-NOVA

Background

This is the day 1 computer lab session for quantitative genetic analysis. The goal of this exercise is to understand mixed model equations (MME) using an interactive MME visualization tool LMMonBoard.

Exercise 1

  1. Select Blank Example from the Data drop-down list.
  2. In Model equation, remove + X2 + X5 so that you have Y1 = Animal + X1.
  3. In the Fixed / Random Effects column, select Animal from the Random Effects (i.i.d.) drop-down list.
  4. Click Update LMM Results on the right upper corner.

Assume that the residual variance \(\sigma^2_e = 1\) and the additive genetic variance \(\sigma^2_g = 1\).

Questions

  • How many fixed effects are in the model? How many levels?
  • How many random effects are in the model? How many levels?
  • Interpret each component of MME.

Exercise 2

  1. In Model equation, add + X2 so that you have Y1 = Animal + X1 + X2.
  2. In the Fixed / Random Effects column, select Animal from the Random Effects (i.i.d.) drop-down list.
  3. Click Update LMM Results on the right upper corner.

Questions

  • How many fixed effects are in the model? How many levels?
  • How many random effects are in the model? How many levels?
  • Interpret each component of MME.

Exercise 3

  1. In the Fixed / Random Effects column, select Animal from the Random Effects (Correlated) drop-down list.
  2. Click Update LMM Results on the right upper corner.

Questions

  • How many fixed effects are in the model? How many levels?
  • How many random effects are in the model? How many levels?
  • Interpret each component of MME.

Exercise 4

  1. In Model equation, add + X5 so that you have Y1 = Animal + X1 + X2 + X5.
  2. In the Fixed / Random Effects column, select Animal from the Random Effects (Correlated) drop-down list.
  3. In the Continuous / Categorical Variables column, select X5 from the Continuous drop-down list.
  4. Click Update LMM Results on the right upper corner.

Questions

  • How many fixed effects are in the model? How many levels?
  • How many random effects are in the model? How many levels?
  • Interpret each component of MME.