Table 3 Predictor composition in PLR, based on chromosomal location.

From: Sexual epigenetics: gender-specific methylation of a gene in the sex determining region of Populus balsamifera

Chromosome

# Features[1]

# Unique features[1]

Rel. cluster prevalence[2]

Chr1

14.1 (1.85)

5.8 (0.87)

9.1

Chr2

3.3 (1.40)

1.9 (0.51)

6.0

Chr3

2.5 (0.85)

1.8 (0.59)

4.5

Chr4

2.4 (0.83)

1.9 (0.58)

3.7

Chr5

2.1 (0.66)

1.6 (0.55)

3.5

Chr6

2.8 (0.90)

2.2 (0.54)

4.6

Chr7

1.2 (0.68)

1.0 (0.44)

3.5

Chr8

1.6 (0.84)

1.0 (0.45)

5.1

Chr9

0.7 (0.37)

0.6 (0.29)

4.3

Chr10

1.5 (0.68)

1.3 (0.49)

3.1

Chr11

2.3 (0.78)

1.8 (0.52)

4.8

Chr12

1.1 (0.55)

0.9 (0.41)

3.7

Chr13

1.7 (0.66)

1.4 (0.48)

4.2

Chr14

1.9 (0.81)

1.3 (0.52)

5.0

Chr15

1.0 (0.63)

0.8 (0.36)

3.4

Chr16

1.1 (0.50)

0.9 (0.40)

3.8

Chr17

1.2 (0.64)

1.0 (0.45)

3.3

Chr18

1.7 (0.61)

1.2 (0.43)

4.8

Chr19

11.1 (0.83)

2.0 (0.43)

37.1

scaffolds

4.2 (1.14)

2.8 (0.72)

5.4

  1. [1] = mean (SD); [2] = average feature prevalence relative to minimal expectance. Individual features can be part of several predictors to account for overlapping pathways. Thus, the total number of features and the number of unique features is given. Relative cluster prevalence indicates the recurrent selection of features for model generation. High values indicate high predictive relevance and stable contribution to model building (K = 5, R = 50).
Back to article page