Table 1.
Evaluation of tetranucleotide sequences for positional disequilibria in different Drosophila promoter datasets.
| Dataset | Element | Highest Peak (location) | SD above | BG-average | SD |
| Ohler et al. [21] | ATAA | 220 (-29) | 5.3 | 65.2 | 29.4 |
| CATG | no peak | - | 18.7 | 4.8 | |
| TATA | 209 (-30) | 5.8 | 49.7 | 27.6 | |
| TCAG | 280 (-2) | 7.6 | 31.3 | 32.7 | |
| TCAT | 104 (-1) | 6.6 | 28.9 | 11.3 | |
| TSS -250..+50 | ATAA | no peak | - | 279.7 | 77.4 |
| CATG | no peak | - | 73.5 | 13.3 | |
| TATA | no peak | - | 213.3 | 69.7 | |
| TCAG | 665 (-1) | 7.7 | 115.5 | 71.6 | |
| TCAT | 268 (-1) | 7.1 | 118.8 | 20.9 | |
| ATG -250..+50 | ATAA | no peak | - | 322.4 | 102.4 |
| CATG | 2806 (-1) | 8.5 | 124.8 | 315.9 | |
| TATA | no peak | - | 224.6 | 79.8 | |
| TCAG | no peak | - | 157.0 | 31.8 | |
| TCAT | 628 (-2) | 8.7 | 157.8 | 54.1 | |
| TSS -1500..+50 | ATAA | 466 (-31) | 17.6 | 196.2 | 15.3 |
| CATG | no peak | - | 105.7 | 10.6 | |
| TATA | 447 (-33) | 19.5 | 157.9 | 14.9 | |
| TCAG | 665 (-1) | 55.1 | 103.6 | 10.2 | |
| TCAT | 268 (-1) | 12.5 | 125.5 | 11.4 | |
| ATG -1500..+50 | ATAA | 567 (-3) | 12.0 | 285.3 | 23.5 |
| CATG | 2806 (-1) | 204.1 | 136.2 | 13.1 | |
| TATA | 357 (-264) | 6.1 | 227.8 | 21.0 | |
| TCAG | 243 (-5) | 9.4 | 137.4 | 11.2 | |
| TCAT | 628 (-2) | 34.9 | 169.1 | 13.1 | |
| TSS -1500..-1 | ATAA | 466 (-26) | 17.6 | 196.1 | 15.3 |
| CATG | no peak | - | 105.7 | 10.6 | |
| TATA | 447 (-29) | 19.6 | 157.9 | 14.8 | |
| TCAG | 226 (-21) | 12.0 | 103.6 | 10.2 | |
| TCAT | 174 (-1041) | 4.2 | 125.5 | 11.5 | |
| ATG -1500..-1 | ATAA | 469 (-196) | 7.8 | 285.2 | 23.5 |
| CATG | no peak | - | 136.2 | 13.1 | |
| TATA | 357 (-263) | 6.2 | 227.7 | 21.0 | |
| TCAG | 243 (-4) | 9.4 | 137.4 | 11.2 | |
| TCAT | no peak | - | 169.1 | 13.1 | |
The hand edited dataset from Ohler et al. [21] was compared with datasets based on annotated promoters that were extracted from genomic sequences derived from the NCBI [59]. The evaluation was conducted using motif distribution curves. Based on previous analysis [20], the significance cutoff frequency for a positional disequilibrium must differ by ≥ 4 SD above the overall background average [see Additional File 8]. no peak: based on our evaluation, no significant peaks could be identified.