Table 4 |
|||||||||
|
The performance of AL-BP-MD-ME-MS on a dataset of shuffled margin sequences a). |
|||||||||
|
nPC |
nSn |
nSp |
|||||||
|
|
|||||||||
|
Margin size (nt) |
100 |
200 |
400 |
100 |
200 |
400 |
100 |
200 |
400 |
|
Original set b) |
0.288 |
0.254 |
0.197 |
0.328 |
0.292 |
0.234 |
0.416 |
0.360 |
0.270 |
|
Shuffled set c) |
0.317 |
0.255 |
0.187 |
0.340 |
0.275 |
0.201 |
0.481 |
0.375 |
0.266 |
|
|
|||||||||
|
a) To the both sides of known sites in the ECRDB61B data set, artificially shuffled sequences with the size of 100, 200, and 400 nt are attached. The statistics of the di-mer nucleotide frequency used to generate the shuffled margin sequences are taken from intergenic regions of the E. coli genome. b) The performance on the original ECRDB61B-100, 200, 400 set. c) The performance on the data set with shuffled margin sequences. |
|||||||||
|
Hu et al. BMC Bioinformatics 2006 7:342 doi:10.1186/1471-2105-7-342 |
|||||||||
