"Seeing the Science: Statistical Graphics for Plant Trait and Genomic Data"

🔍 1. Descriptive Statistics & Trait Analysis

Used to summarize and explore phenotypic data (e.g., yield, plant height, flowering time):

Boxplots – Compare trait distributions across genotypes, treatments, or locations.

Histograms – Show the frequency distribution of traits.

Violin plots – Similar to boxplots but show kernel density.

Bar charts – Compare means of traits (often with error bars).

Scatter plots – Show relationships between two continuous traits.

Dot plots – Visualize trait values for each genotype.

🌾 2. Genotype × Environment Interaction (G×E)

To evaluate how genotypes perform across environments:

GGE Biplot (Genotype + Genotype × Environment) – Visualize performance and stability of genotypes.

AMMI Biplot (Additive Main Effects and Multiplicative Interaction) – Partition and display G×E interaction.

Interaction plots – Line plots showing performance across environments.

Heatmaps of trait performance across environments – Easily spot high- and low-performing combinations.

🧬 3. Genetic Diversity & Population Structure

Helpful when assessing variability, clustering, and population stratification:

• Principal Component Analysis (PCA) plots – Show genetic variation among genotypes.
• Dendrograms (Hierarchical Clustering) – Group genotypes based on genetic similarity.
• STRUCTURE bar plots – Visualize population structure (from STRUCTURE or ADMIXTURE outputs).
• Discriminant Analysis of Principal Components (DAPC) – Genetic clustering visualization.
• Multidimensional Scaling (MDS) plots – Similar to PCA for distance-based visualization.

🧪 4. QTL Mapping / GWAS (Genetic Association Studies)

Used to identify genetic markers associated with traits:

• Manhattan plots – Show significant associations between markers and traits.
• QQ plots (Quantile-Quantile plots) – Assess false positive rates in GWAS.
• Linkage disequilibrium (LD) heatmaps – Visualize the degree of linkage between markers.
• Circos plots – Integrate QTL/GWAS results across chromosomes.

📈 5. Time-Series / Growth Analysis

For analyzing growth over time or developmental stages:

• Growth curves / line plots – Trait progression over time.
• Area plots – Useful for cumulative growth or production.
• Spaghetti plots – Overlay individual genotype trends.

📊 6. Heritability & Variance Component Analysis

To assess trait control and genetic contribution:

• Variance partitioning bar plots – Show % contribution of genetic, environmental, and error variances.
• Error bar plots – Show confidence intervals around heritability estimates.
• Pie charts / stacked bar plots – Visual breakdown of variance components.

🧪 7. Experimental Design & Field Layout

Useful for visualizing trial setup and data collection:

• Field layout maps / grid plots – Show design (e.g., RCBD, alpha lattice).
• Heatmaps of plot values – Spot spatial variability in fields.
• 3D surface plots – Depict field variability in traits like yield.

🔄 8. Correlation & Multivariate Analysis

To explore relationships between multiple traits or variables:

• Correlation matrix heatmaps – Trait-to-trait relationships.
• Pairwise scatterplot matrix – Multiple scatter plots with correlation overlays.
• Biplots (PCA or PLS) – Multivariate representation of genotypes and traits.
• Cluster heatmaps – Combine trait clustering and visualization.

🧬 9. Genomic Selection

Visualizations used in prediction modeling and model evaluation:

• Predicted vs. Observed scatter plots – Assess accuracy of genomic predictions.
• Cross-validation performance plots – Compare models using RMSE, R², etc.
• Feature importance bar plots – Show most important markers or traits in prediction.

📌 Bonus: Specialized Visualizations

• Sankey diagrams – Flow of genetic material across generations (e.g., pedigree tracking).
• Alluvial plots – Changes in group membership or trait classification over time.
• UpSet plots – Better alternative to Venn diagrams for multiple set intersections (e.g., shared QTLs).

---