Skip to main content

Table 4 Mean and standard deviation of the Pearson correlation for the SVR models constructed from subsets of SNPs selected by adjusted p-values of the simulated database 1.

From: SNPs selection using support vector regression and genetic algorithms in GWAS

Group

SNP selection with adjusted p-values*

# SNPs

Linear**

RBF**

PUK**

1

< 10-9

1a

0.36(0.17)

0.37(0.17)

0.45(0.17)

2

< 10-8

1a

0.36(0.17)

0.37(0.17)

0.45(0.17)

3

< 10-7

1a

0.36(0.17)

0.37(0.17)

0.45(0.17)

4

< 10-6

1a

0.36(0.17)

0.37(0.17)

0.45(0.17)

5

< 10-5

1a

0.36(0.17)

0.37(0.17)

0.45(0.17)

6

< 10-4

1a

0.36(0.17)

0.37(0.17)

0.45(0.17)

7

< 10-3

1a

0.36(0.17)

0.37(0.17)

0.45(0.17)

8

< 10-2

1a

0.36(0.17)

0.37(0.17)

0.45(0.17)

9

< 10-1

3b

0.49(0.14)

0.55(0.14)

0.56(0.13)

10

< 0.20

3b

0.49(0.14)

0.55(0.14)

0.56(0.13)

11

< 0.30

3b

0.49(0.14)

0.55(0.14)

0.56(0.13)

12

< 0.40

4 c

0.53(0.15)

0.56(0.14)

0.78(0.16)

13

< 0.50

4c

0.53(0.15)

0.56(0.14)

0.78(0.16)

14

< 0.60

4c

0.53(0.15)

0.56(0.14)

0.78(0.16)

15

< 0.70

5d

0.53(0.15)

0.56(0.14)

0.75(0.16)

16

< 0.80

6e

0.55(0.14)

0.58(0.14)

0.71(0.16)

17

< 0.90

6e

0.55(0.14)

0.58(0.14)

0.71(0.16)

18

-

7f

0.65(0.09)

0.65(0.09)

0.93(0.02)

  1. *Adjusted p-values of the Spearman correlation with Bonferroni correction.
  2. ** Mean and standard deviation of the Pearson correlation in 10-fold cross validation.
  3. (a) SNP1. (b) SNP1, 10 and 60. (c) SNP1, 10, 50 and 60. (d) SNP1, 10, 50, 60 and 874. (e) SNP1, 10, 50, 60, 121 and 874. (f) Correct model with 7 relevant SNPs and without redundant SNPs.
  4. For each group defined after the first selection, a model SVR was constructed, evaluated for each kernel, through the Pearson correlation coefficient in 10-fold cross validation. The bold line indicates the best model from the first selection for PUK.