Skip to main content

Table 1 Evaluation of nSVM results on the spiking dataset using different parameters

From: PatternLab for proteomics: a tool for differential shotgun proteomics

Elitism

Mutation

Islands

No. Feat.

Avg. No. Subst.

Time

PHS2

ALB

CAH

ITRA

0

2

0

6

290

49 ± 9

1

3

5

4

0

2

250

4

380

50 ± 8

1

2

4

3

0

3

0

5

400

54 ± 6

2

1

3

5

0

3

250

4

415

51 ± 9

1

3

5

6

1

1

250

4

280

49 ± 7

4

1

2

3

1

2

0

6

360

56 ± 7

1

2

4

3

1

2

250

4

410

59 ± 3

1

3

4

2

1

3

0

5

416

54 ± 5

5

3

4

7

1

3

250

6

423

52 ± 7

1

6

3

2

  1. nSVM was executed 20 times for each of the 8 specified conditions. PHS2, ALB, CAH, and ITRA stand for the spiked protein markers, and the numbers in the respective columns indicate the ranking. The Elitism column stands for how many individuals of the population were allowed to remain untouched for the following generation. The Islands column indicates how many seconds were required before a migration could even occur; a zero indicates that the island model was not applied. The No. Feat. column indicates what nSVM suggested as the minimum set of optimal features. The Avg. No. Subst. column indicates how many times the fittest individual was substituted. The Time column indicates the average time and standard deviation of 1 nSVM run. These results were obtained for scenario 1 as described in the Suggestion of when to apply nSVM section. Due to the stochastic nature of the method, results may vary.