Table 5 |
||||||
|
Complementarity within the six unique functional site prediction schemes.1 |
||||||
|
PM |
FPE |
SC85 |
Rate4Site |
ET |
SDPpred |
|
|
|
||||||
|
PM |
--- |
--- |
--- |
--- |
--- |
--- |
|
FPE |
0.62 |
--- |
--- |
--- |
--- |
--- |
|
SC85 |
0.62 |
0.69 |
--- |
--- |
--- |
--- |
|
Rate4Site |
0.66 |
0.49 |
0.59 |
--- |
--- |
--- |
|
ET |
0.69 |
0.64 |
0.62 |
0.62 |
--- |
--- |
|
SDPpred |
0.80 |
0.66 |
0.86 |
0.71 |
0.64 |
--- |
|
|
||||||
|
1 Complementarity is calculated as the Euclidean distance between each vector pair describing the results within Table 2, meaning each vector has 34 dimensions. Each dimension is simply a binary possibility (1 = predicted functional site; 0 = not predicted functional site). The values reported are normalized such that two completely orthogonal vectors will have a distance of one; two vectors that are absolutely the same will have a distance of 0.00, whereas two completely orthogonal vectors will have a distance of 1.00. The average value and standard deviation across the matrix are 0.66 and 0.09, respectively, which corresponds to 15.0 and 4.1 differences. |
||||||
|
Livesay et al. BMC Bioinformatics 2007 8:397 doi:10.1186/1471-2105-8-397 |
||||||
