Article banner region

Linear Models in Animal Breeding

  • Photo: 
    NMBU

NOVA PhD course of 3 ECTS, organised by Susanne Eriksson, Swedish University of Agricutlural Sciences.
Preliminary course dates and location: 17-21 Jun. 2018, in Orsa Grönklitt, Sweden.

Course Description
This course in linear models in animal breeding will be based on theoretical lectures alternating with practical computer exercises, mainly in R. The course will include pre-course exercises and literature, and a pre-course online meeting for those who want to ask questions about the exercises.

Content
The course will include sections on linear regression, diagnostic tools for linear regression, linear regression using weighted least squares, maximum likelihood, a brief introduction to generalized linear models, linear mixed models, introduction to GWAS models, and genomic selection using gBLUP, short overviews of dispersion modelling, GLMM, and Bayesian alternatives.

Programme Outline
Lectures and computer exercises will be mixed throughout the day.

  • Day 1
    • Introduction to the course
    • Linear regression
    • Diagnostic tools for linear regression
    • R exercises

  • Day 2
    • Linear regression using weighted least squares
    • Maximum likelihood
    • A brief introduction to generalized linear models
    • R exercises

  • Day 3
    • Linear mixed models – independent random effects
    • Linear mixed models – animal model
    • R exercises

  • Day 4
    • Introduction to GWAS models
    • Introduction to genomic selection using gBLUP
    • R exercises

  • Day 5
    • Linear mixed models
    • Questions and clarifications
    • R exercise
    • Examples of more advanced models
    • Summary of the course

Pre-/Post-Campus Assignments
Three weeks before the course starts some pre-course exercises will be given, including exercises in R and some questions on linear regression and statistical theory. Students are expected to have completed a majority of these exercises before arriving to the course. One week before the course start there will be an online meeting for those who want to ask questions on the pre-course exercises.

Learning Outcomes
After completing the course the students shall be able to:

  • Choose appropriate models for different analyses.
  • Use diagnostic tools in R for linear regression.
  • Perform simple genome wide association analysis.
  • Describe and explain the basic principles of genomic selection.
  • Find and use relevant information on more complex situations and analyses when needed.

Evaluation Elements
Completed course exercises.

Pedagogical Approach
There will be mainly lectures and computer exercises during the course, and pre-course self-studies of literature and computer exercises with possibility to ask questions. During the course the students are expected to be active, and to discuss the topics and the exercises with each other and with the teachers.

Estimated Workload

  • seminar: 6 hours
  • lecture: 17 hours
  • independent work: 50 hours (including pre-course reading literature, pre-course exercises and discussion via Skype) 
  • other: computer exercise approx 17 hours

Prerequisite Knowledge
Doctoral student experience or similar, with knowledge in animal science or veterinary medicine, or participant in a residency program in veterinary science.

Admission
Admission for NOVA courses is handled by the course organiser/ the NOVA member institution organising the course. Please see the links in the margin for more information.

Published 13. June 2017 - 10:50 - Updated 13. June 2017 - 10:50

NOVA University Network

E-mail:
secretariat@nova-university.org
 

Postal address:
Ingrid Aksnes Hjetland
NOVA Secretariat
Norwegian University of Life Sciences
PO BOX 5003
NO-1432 Aas
NORWAY