Build ANOVA Heatmap - hisl6802/ClusteringToolbox GitHub Wiki

Building an ANOVA Heatmap

Upon looking for the dysregulated metabolites in an untargeted metabolomics study with three experimental conditions (i.e., injured, sham-injured, and contralateral-to-injured) an ANOVA and a hierarchical clustering are often performed. A common visualization after performing these analyses is an ANOVA Heatmap where the significant metabolites identified through ANOVA are clustered using a agglomerative hierarchical clustering, allowing the user to visualize which metabolites are dysregulated in the "same" manner (i.e., up and/or down regulated) with a disease condition.

Input file (-log10(p) may show up as $NAME? in excel)

	f.value	p.value	-log10(p)	FDR	Post-hoc tests
685.85	32.967	3.4E-8	7.4679	5.17E-5	CtInj - Sinj; Inj - Sinj
967.82	27.656	1.9E-7	6.7206	1.45E-4	CtInj - Sinj; Inj - Sinj
...	...	...	...	...	...

Output

The output will be a heatmap of the metabolites which cross a wanted threshold.