Evaluation of phylogenetic footprint discovery for predicting bacterial cis-regulatory elements and revealing their evolution

Table 4 Significance of the impact of individual parameters on the accuracy of discovered motifs

Parameter	Better	Worse	N	Sup	Inf	Equal	P-value
Background model	taxfreq	monad	20	20	0	0	9.54E-07
Dyad filtering	filtered	not filtered	20	20	0	0	4.78E-05
Taxon	Gammaproteobacteria	Bacteria	8	8	0	0	0.004
Taxon	Proteobacteria	Bacteria	8	8	0	0	0.004
Taxon	Enterobacteriales	Escherichia	8	6	2	0	0.020
Taxon	Gammaproteobacteria	Escherichia	8	6	2	0	0.020
Taxon	Gammaproteobacteria	Proteobacteria	8	6	2	0	0.034
Taxon	Proteobacteria	Escherichia	8	6	2	0	0.070
Taxon	Enterobacteriales	Bacteria	8	6	2	0	0.074
Promoter	orthologs	operon leader	20	13	5	2	0.107
Taxon	Bacteria	Escherichia	8	4	4	0	0.191
Taxon	Gammaproteobacteria	Enterobacteriales	8	5	3	0	0.273
Taxon	Enterobacteriales	Proteobacteria	8	4	4	0	0.528

Parameters are sorted according to the significance of their impact on the geometric accuracy. The best parameter value of each pair is in the left column. N: number of parametric combinations. Sup, inf, equal: number of pairs where the left value gives better, wose or identical results than the right value, respectively. The P-value is estimated with a Wilcoxon paired test.

ISSN: 1471-2105