INFLUENCE_ANOVA - Anova of a variable characterising pairs reactions
p = influence_anova(D, n_per, fdr, ind_intervention)
Anova of a variable characterising pairs reactions
(e.g., all second-order control coefficients for a single objective)
Arguments:
D: data tensor (size nr x nr x n_sim x 2) where
nr: # reactions (corresponds to 'reaction' vector)
n_sim: # number of MC samples
2: two qualitative conditions to be compared (e.g. different fluxes)
fdr: false discovery rate
ind_intervention: reaction indices to be selected for the analysis (optional)
The results are returned in a structure 'p' with fields
p.mean_total Mean value for all samples (total mean)
p.mean_a Mean value for 1st variant
p.mean_b Mean value for 2nd variant
p.mean_delta Difference between values for both variants
p.p_value_mean_total p value of [mean value for all samples ~= total mean]
p.p_value_mean_a p value of [mean value for 1st variant ~= total mean]
p.p_value_mean_b p value of [mean value for 2nd variant ~= total mean]
p.p_value_mean_delta p value of [difference between values for both variants
~= difference between values from shuffled variants]
p.mean_total_significant Significant elements for mean value for all samples
p.mean_a_significant Significant elements for mean value for 1st variant
p.mean_b_significant Significant elements for mean value for 2nd variant
p.mean_delta_significant Significant elements for difference between values for both variants
Entries: 1 (for significantly high), -1 (for significantly low)
p.n_mean_total_significant # significant elements for mean value for all samples
p.n_mean_a_significant # significant elements for mean value for 1st variant
p.n_mean_b_significant # significant elements for mean value for 2nd variant
p.n_mean_delta_significant # significant elements for difference between values for both variants