|
P > 0.5
|
P > 0.5 **
|
P > 0.8
|
P > 0.9
|
---|
TAIR data set
|
12/55 (22%)
|
6/55 (11%)
|
7/55 (13%)
|
5/55 (9%)
|
2/55 (4%) ¶,
|
2/55 (4%)
|
1/55 (2%)
|
1/55 (2%)
|
Matsui2008
|
71/1365 (5%) ¶
|
--
|
46/1365 (3%)
|
32/1365 (2%)
|
Okamoto2010
|
22/485 (5%) ¶
|
--
|
13/485 (3%)
|
10/485 (2%)
|
- * The fractions are of the totals that have significant sequence conservation. For the TAIR data, the first row is for RNAz calculations across the whole NOLP sequence. For non-TAIR data and for the second row for the TAIR data, these are cases that overlap CNSs (conserved non-coding sequences) identified in Haudry, et al. [21].
- ** Removing non-significantly conserved cases after Holm-Bonferroni correction.
- ¶ The total numbers of conserved are significantly enriched compared to randomly sampled near-gene DNA (P = 0.046 for the TAIR set, P < 0.00001 for the other two sets, normal statistics). To assess this, for each of the three actual data sets listed, 500 samples of near-gene DNA of the same distribution of sizes and position relative to neighbour genes as the actual set were generated (as described in the
Methods
).