Estimating enrichment of repetitive elements from high-throughput sequence data

Table 2 Agreement between enrichment estimates based on reads mapping to canonical and instance sequences

	H3K27me3	H3K36me3	H3K4me3	H3K9me3	H4K20me3	RNA Pol II
Fraction of repeat types with intersecting canonical and instance-based CIs	0.84	0.93	0.84	0.93	0.92	0.63
Average ratio of canonical and instance-based CI sizes	25.6	24.5	25.9	22.5	22.1	20.0

Using mES data for different chromatin marks, the table shows the fraction of repeat types for which estimates based on canonical and instance read counts were statistically consistent. Only repeat types for which the canonical confidence interval (CI) did not include 0 (that is, significant enrichment was detected) were considered in the analysis. The canonical and instance-based estimates are typically consistent for 70% of the repeat types; however, instance-based estimates provide, on average, much smaller confidence intervals.

ISSN: 1474-760X