TITULO: PREDICTION OF COMPLEX TRAITS

IMPARTIDO POR: DANIEL GIANOLA

LENGUA: Impartido en castellano con diapositivas en inglés

FECHA: 22 a 26 de Mayo de 2017

LUGAR: Edificio del Departamento de Ciencia Animal (Edificio 7G). Universidad Politécnica de Valencia. Aula 1 (Planta baja) . Pinchar aquí para ver el plano de situación

HORARIO:

LUNES a VIERNES    9:30 a 14:00

MATRICULA: Se paga un precio de 50 euros por la gestión y los diplomas. Acceder a la siguiente dirección y seguir las instrucciones:

https://www.cfp.upv.es/formacion-permanente/cursos/prediccion-de-caracteres-complejos_idiomaes-cid52548.html

 

 

PROGRAMA


1. Introduction. Molecular markers and prediction. Predictive inference. Cross-validation. Overview of some penalized methods.
 

2. Review of least-squares, maximum likelihood and best linear unbiased prediction.
 

3. GWAS (genome-wide association study) and pitfalls.
 

4. Review of Bayesian inference, MCMC and Bayesian regression. Bayesian predictive distributions
 

5. Challenges from complexity: over-parameterization, instrumental models, errors in gene action specification.
 

6. Genomic BLUP and genomic studentized prediction (GSTUP). The Bayesian alphabet (Bayes A, B, C, Bayesian Lasso, Bayes R).
 

7. The problem of dealing with gene-gene-gene-....-gene interactions.
 

8. Introduction to non-parametric regression: kernel methods and neural networks.
 

9. Estimating distributions of prediction errors via re-sampling.
 

PRESENTACIONES

 

FILE 1-INTRODUCTION.pdf

 

FILE 2 A- OVERVIEW OF LEAST SQUARES ML BLUP PART 1.pdf

 

FILE 2 B- OVERVIEW OF LEAST SQUARES ML BLUP PART 2.pdf

 

FILE 3- OVERVIEW OF BAYESIAN INFERENCE.pdf

 

FILE 4- GENOMIC BLUP AND BAYESIAN GENOMIC BLUP.pdf

 

FILE 5- THE BAYESIAN ALPHABET

 

FILE 6 A - RKHS REGRESSION