Skip to main content

Table 3 Robustness of correlation to variations in window size.

From: Correlation of microsynteny conservation and disease gene distribution in mammalian genomes

Window size

20 Mb

10 Mb

5 Mb

2 Mb

1 Mb

Actual annotation

R = 0.40,

p < 4 × 10-4

R = 0.293,

p < 1 × 10-10

R = 0.256,

p < 1 × 10-10

R = 0.157,

p < 1 × 10-10

R = 0.154,

p < 1 × 10-10

Randomized annotation

R = 0.02,

p = 0.58

R = 0.018,

p = 0.59

R = 0.058,

p = 0.11

R = 0.046,

p = 0.45

R = 0.035,

p = 0.01

Actual annotation, empty windows removed

   

R = 0.11

p < 1 × 10-10

R = 0.11,

p < 1 × 10-10

Randomized annotation, empty windows removed

   

R = 0.012,

p = 0.40

R = 0.020,

p = 0.83

  1. Adjusting the window size does not eliminate the correlation between regions of conserved microsynteny and regions of high density of disease gene orthlogs. Randomization of gene annotations does eliminate the correlation at all window sizes except 1 Mb. However, when windows with no annotated genes are omitted from the analysis (bottom two rows), the correlation for actual annotations is improved at both 2 Mb and 1 Mb window sizes, and is eliminated for random annotations. As the genome does not have any windows lacking gene annotation for windows of 20 Mb, 10 Mb, or 5 Mb, the correlation values could not be recalculated to omit annotation-free windows at those window sizes. R = Pearson's R.
Back to article page

BMC Genomics

ISSN: 1471-2164

Contact us