Skip to main content

Table 2 Motif mapping and co-expression statistics

From: The orientation of transcription factor binding site motifs in gene promoter regions: does it matter?

A) Up-stream sequence interval [bp] B) Motifs C) Presence/ absence statistic D) Mapping orientation statistic, D1/D2/D3 E) Orientation and presence/ absence effect (% of C) F) Set E with PE-filter G) Motifs matching criteria E) and, indicated by (*), F)
−500, −1 264 112 (42.4 %) 39 (14.7 %)#/
27 (10.2 %)/
16 (6.1 %)
7 (6.2 % of 112) 6 M0007_1.01*, M0252_1.01*, M0576_1.01, M1180_1.01*, ABRE-like_binding_site_motif*,
−250, −1 264 135 (51.1 %) 65 (24.6 %)/
32 (12.1 %)/
13 (9.6 % of 135) NA M0007_1.01, M0023_1.01, M0036_1.01,
M0078_1.01, M0264_1.01, M0516_1.01,
M0576_1.01, M0578_1.01, M0770_1.01,
Box_II_promoter_motif, LEAFYATAG
−100, −1 247 140 (56.7 %) 57 (23.1 %)#/
23 (9.3 %)/
11 (7.8 % of 140) NA M0014_1.01, M0016_1.01, M0019_1.01, M0031_1.01, M0078_1.01, M0081_1.01, M1180_1.01, Bellringer/replumless/ pennywise_BS1_IN_AG, Box_II_promoter motif, LTRECOREATCOR15, MYB_binding_site motif
−500, −51 260 100 (38.4 %) 28 (10.8 %)/
18 (6.9 %) /
5 (5 % of 100) NA M0007_1.01, M0252_1.01, ACGTABREMOTIFA2OSEM, GADOWNAT, GAREAT
Core promoter motifs
−50, −1 10 6 (60 %) 5 (50 %) /
1 (TATA-box, 10 %) /
0 (0 % of 4) NA No significant motifs according to filter criteria E detected
  1. Motif mapping and co-expression analysis results. Table columns list information on A) the interval of the considered upstream regions, B) the number of considered motifs with valid observations (Note that the number of considered motifs (column B) is less than the reported set size of 293 motifs of initially compiled motifs (see Methods) and differs between test settings as for the motifs that dropped out either i) mapping counts were insufficient, or ii) no probes were present on the ATH1 chip for the respective downstream gene, and we report the results of motifs with complete information across all tests (mapping, expression, and, in the case of 500 bp upstream region, position entropy) only.), C) Number (percentage) of motifs with significant co-expression differences between genes containing the genes upstream regardless of direction compared to genes not containing the motif at all (neither in forward nor reverse-complement orientation) with thresholds pr_diff < 0.05 and Cohen’s d > 0.01. D) Motif mapping statistics with D1 indicating the number of motifs with significant orientation preference (porient < 0.05), D2 - subset of D1 meeting also the criteria of significant co-expression differences (pr_diff < 0.05) with higher intra-set correlations in the set corresponding to the preferred mapping orientation, and, in addition (D3), lowered positional entropy (PE) in the preferred orientation. As no positional entropies were computed for the shorter upstream intervals of length 250 bp and 100 bp, D3 is not provided for those sets. E) Filter criteria D2 applied only to the subset of motifs with evidence of significant presence/absence effect (column B) (Note that the multiple testing correction was adjusted accordingly.) F) Subset of E that also exhibit lowered positional entropy (PE) in the preferred orientation (Filter criteria D3, applied to upstream regions of length 500 bp only as positional preferences lose their meaning for smaller considered sequence intervals). G) Actual motif names fulfilling filter criteria E, and if indicated by asterisks, F. Underlined values denote counts and percentages significantly different from random expectation (p < 0.01) based on Fisher’s exact tests with randomly expected counts determined from mapping statistics and expression analyses obtained for sets of random motifs with compositions based on the considered upstream regional interval as well as upstream motifs (Additional file 2: Table S1; see Methods); i.e., found significant relative to both randomized sets R1 and R2. § - significantly different relative to upstream-composition-based randomization (R1 Additional file 2: Table S1) (This case was not observed), # - significantly different relative to motif-composition-based randomization (R2 Additional file 2: Table S1) with significance judged after correcting for multiple testing