Fungal taxonomic abundance heatmaps

Z-score heatmaps of fungal abundance across phylum, family, genus, and species levels by study group.

1 Research question

What patterns of fungal taxonomic abundance are visible across study groups (Non-IBD, Active IBD, Quiescent)? This is a descriptive visualization — no hypothesis test or p-value is computed here.

2 Data

• Response: Log₁₀-transformed, z-score normalized relative abundance per taxon.
• Predictors: Study group annotation per sample column.
• Unit of analysis: Sample × taxon.
• Input files: data/intermediate/taxa_long_list.rds, data/intermediate/meta_data.rds (from make mycobiome).
• Filtering: Taxa with zero variance, zero total abundance, or prevalence < 10 % of samples are excluded per taxonomic level.

3 Methods

Item	Choice
Transformation	log₁₀(abundance + 1×10⁻⁶) per taxon
Scaling	Z-score across samples per taxon (row-wise)
Clustering	Hierarchical (ComplexHeatmap defaults) on both rows and columns
Annotation	Study group colour bar (top)
Package	ComplexHeatmap
Multiple testing	Not applicable — no statistical test performed

Figures generated by src/mycobiome/05_heatmap_analysis.R via make mycobiome.

4 Figures

4.1 Phylum

Z-score heatmap of fungal phyla. Columns are samples, rows are phyla passing prevalence filter. Top annotation shows study group.

4.2 Family

Z-score heatmap of fungal families.

4.3 Genus

Z-score heatmap of fungal genera.

4.4 Species

Z-score heatmap of fungal species.

5 Interpretation

Heatmaps show relative abundance patterns across samples grouped by IBD status. Clustering of samples and taxa is driven by abundance co-variation, not by group labels — any apparent group separation is observational. Formal tests of group differences are in the PERMANOVA posts (make stats-permanova).