Randomized pairs were selected from 27 910 non-NAT genes (including lncRNAs) with 10 000 repeats.
Randomized pairs were selected from single copy genes with 10000 repeats.
For trans-NAT gene pairs, similar results hold (Wilcoxon's rank-sum test: P < 0.05 for five cases), except at CHH sites, where trans-NAT gene pairs and randomized pairs show no difference in their DNA methylation distance.
To further compare the epigenetic patterns between NAT pairs and randomized pairs in each group, two sets genes were randomly selected as the control sets from non-NAT genes, which have the same gene number with the cis-NATs and trans-NATs groups.
To minimise any potential bias the composition of both sets was randomized after pairing miRNA expression indexes with their respective mRNA expression values.
To test this hypothesis, we compared the distance of HM code between duplicate pair and randomized singleton pair.
To rule out this possibility, we chose duplicate pairs and randomized singleton pairs where both copies are belonging to different chromosomes, and observed the similar result to Figure 1A (see Additional file 1: Figure S1).
To calculate random expectations, we broke the pairing between alternative promoters of the same gene, or neighboring unique promoters, and randomized the pairing 10,000 times for each type.
Following this argument, we classified all yeast gene pairs (duplicate and randomized singleton pairs) under study into two groups: they are located in the same chromosome, or different chromosomes, and tested the relationship between histone modification pattern and chromosome condition.
