Skip to main content

Table 4 Guide strand position specific base composition (Method 1) for training RBF-epsilon regression SVM model

From: Improving model predictions for RNA interference activities that use support vector machine regression by combining and filtering features

  

train579

train579

train579

  

test579

test2431

test579 10 × cross val

t- test

FN579

R

MSE

R

MSE

R

MSE

0

84(84)

0.716

0.048

0.484

0.054

0.562

0.079

1

45(45.9)

0.718

0.048

0.483

0.056

0.541

0.081

2

22(21)

0.645

0.057

0.467

0.058

0.449

0.091

3

8(7.1)

0.489

0.075

0.353

0.055

0.419

0.092

4

4(3.7)

0.418

0.082

0.350

0.052

0.424

0.092

5

3(2.2)

0.397

0.083

0.334

0.051

0.363

0.097

6

2(2)

0.327

0.089

0.304

0.049

0.340

0.099

7

1(0.2)

0.281

0.093

0.176

0.053

-

-

8

0(0)

-

-

-

-

-

-

9

0(0)

-

-

-

-

-

-

  1. Models trained on dataset579 and testing performed with dataset579, dataset2431 and 10 × cross validation on dataset579, features removed by increasing stringency of t-test of individual feature to activity from dataset579.
  2. Feature numbers in parentheses are the average number of features in cross validations.
  3. Entries in bold are column maximums and are simply visual landmarks.