Table 1. Statistics across all populations for the single most abundant isomiR produced from each precursor arm.

The most abundant isomiR's identity was determined by analyzing the entire collection of all unique datasets that passed quality control (452 datasets, one for each of 452 individuals; no technical replicates were included). At each locus, we also generated the “conservative” estimate (4th row) by requiring that the most abundant isomiR have higher expression than the miRBase reference by at least 0.585 log2 units.

	Number	Percentage
MiRNA precursor arms considered	194
MiRBase miRNA sequence not among the 95-th percentile	60	31%
Top isomiR not in agreement with miRBase	91	47%
Top isomiR (conservative) not in agreement with miRBase	84	43%
Top isomiR disagrees with miRBase at 5' end	12	6%
Top isomiR disagrees with miRBase at 5' end only	3	2%
Top isomiR disagrees with miRBase at 3' end	88	45%
Top isomiR disagrees with miRBase at 3' end only	79	41%
Top isomiR disagrees with miRBase at both ends	9	5%