PowerPoint Presentations

Biplot Analysis Workshop Outline

ã Weikai Yan, August 2003.

1.    Introduction

1.1.          Data are expensive and precious

1.2.          Data are not used in full

1.3.          Why Biplot

2.    Theory of biplot

2.1.          Biplot display of a Rank-two matrix

2.2.          The reverse: Singular value decomposition

2.3.          Types of multiplicative models

2.4.          Types of data transformation

2.5.          Singular value partitioning methods

2.6.          Physical scales and visualization of a biplot

2.7.          Questions to ask before trying to interpret a biplot

2.8.          Interpretation of a biplot

2.8.1.                   relationship among row factors

2.8.1.1.                          the angle between two row factors

2.8.1.2.                          the length of the vectors

2.8.2.                   relationship among column factors

2.8.2.1.                          the angle between two column factors

2.8.2.2.                          the length of the vectors

2.8.3.                   interaction between row factors and column factors

2.8.3.1.                          the angle between a row and a column

2.8.3.2.                          the length of the vectors

 

3.    GGEbiplot Software

3.1.          File

3.2.          View

3.3.          Four-way data options

3.4.          Biplot analysis tools

3.5.          Multivariate analyses

3.6.          Format

3.7.          Models

3.8.          Biplots

3.9.          Singular value partitioning

3.10.       Data Manipulations

3.11.       Accessories

3.11.1.            Data Plotting

3.11.2.            Numerical output

3.11.3.            Conventional analysis

 

4.    Biplot analysis of genotype by environment data

4.1.          GE interaction and mega-environment classification

4.1.1.                   Clustering of test environments

4.1.2.                   Which-won-where

4.1.3.                   Repeatability (across years)

4.2.          Cultivar evaluation

4.2.1.                   The relative performance of a cultivar in different environments

4.2.2.                   Comparison between two cultivars in individual environments

4.2.3.                   Comparison among three cultivars in individual environments

4.2.4.                   Cultivar ranking in a single environment

4.2.5.                   Cultivar ranking in two environments

4.2.6.                   Cultivar ranking across all environments (mean)

4.2.7.                   Cultivar ranking based on mean and stability

4.2.8.                   (mean and stability deviation)

4.2.9.                   (Yield relative the maximum)

4.2.10.                 Generally adapted cultivars

4.2.11.                 Specifically adapted cultivars

4.2.12.                 Lack-of-response cultivars

4.2.13.                 Cultivars poorest in everywhere: are they poorest in everything? (ref: genotype by trait biplot)

4.3.          Test environment evaluation

4.3.1.                   Representativeness/effectiveness

4.3.2.                   Discriminating ability/informative or not

4.3.3.                   Uniqueness/redundancy

4.3.4.                   Ideal test environments

4.3.5.                   minimum number of test environments

4.4.          Understanding GE interaction

4.4.1.                   Genotypic factors (QTL, simply inherited traits)

4.4.2.                   Environmental factors

5.    Biplot analysis of genotype by trait data

5.1.          Cultivar/parent evaluation

5.2.          Trait relations

5.2.1.                   Trait for indirect selection?

5.2.2.                   Redundant traits?

5.2.3.                   Linear model of an important trait

5.2.4.                   Systems understanding of your plant

6.    Extension of two-way data analysis

6.1.          “G-E-T” three-way data analysis

6.1.1.                   GGE biplot for each trait

6.1.2.                   Genotype by trait biplot in each environment

6.1.3.                   Genotype by trait biplot in some environments

6.1.4.                   G by T table based on genetic data

6.1.5.                   G by T table based on environmental data

6.1.6.                   G by T table based on phenotypic data

6.1.7.                   G by ET for genotype classification

6.2.          Another dimension: “Y-L-G-T” four-way data analysis

6.2.1.                   many times more informative than a three-way dataset

7.    QTL mapping based on phenotypic data from a single environment

7.1.          Markers significantly associated with the target trait

7.2.          Linkage relation among selected markers

7.3.          Multiple regression to identify QTL

7.3.1.                   Markers most closely linked to the QTL (position)

7.3.2.                   Effect of each QTL

8.    Biplot analysis of QTL by environment interaction

8.1.             MET is Mandatory

8.2.             QTL mapping using MET has been problematic

8.3.             Marker-by-environment two-way table

8.4.             QQE biplot

8.5.             Position and effect of QTL

8.6.             QTL by environment interaction

8.7.             Mega-environment identification based QEI

8.8.             Interpreting GEI using QTL and traits

9.    Biplot analysis and multiple regression

10.Biplot analysis of other types of two-way data

10.1.       Diallel data

10.1.1.                 General combining ability

10.1.2.                 Specific combining ability

10.1.3.                 Best entries

10.1.4.                 Best testers

10.1.5.                 Best crosses

10.1.6.                 Heterotic groups

10.1.7.                 Genetic basis

10.2.       Host by pathogen two-way data

10.2.1.                 Race specificity?

10.2.2.                 Vertical vs. horizontal resistance

10.2.3.                 Resistance gene distribution

10.2.4.                 Breeding strategies

10.3.       Environment by factor two-way data