Table 1.

Main characteristics of the selected enrichment methods. Detailed explanations of these equations are provided in the Supplementary Text S1

Method	Class	Statistics	P value assessment	Descriptions
ORA	ORA-based	Hypergeometric		Features and pathways are treated equally; only significant features are used; arbitrary selection for the significant features
Mummichog	ORA-based	FET, EASE		Same as above; using permutation to address the dependence assumption
GSEA	FCS	Kolmogorov–Smirnov-like statistic		Small changes are included; features and pathways are treated equally; summarize the differential information to a rank list
FELLA	TP-based	Diffusion and PageRank algorithms		Combined topology information with ORA; built-in weighted network consisting of genes, enzymes, and metabolites; dependent on the annotated network relationships