Figure 2. Independent evaluation of miRNAs and miRNA-like (mi/milRNA) annotations using sRNA-seq data.
(A) Number of hairpin loci in which the reported mature mi/milRNA sequence is the most-abundant sequence (MAS) (orange), another sequence in the locus is the MAS (pink), the cited sequence is not present in the locus (light gray), or the reported locus has no aligned reads (dark gray). (B) Abundance (aligned reads per million) and strandedness for mi/milRNA loci. Each dot represents a locus, and the colors represent whether the locus passes the 0.5 reads per million cutoff of expression (pass rpm, yellow), has at least 80% of the aligned reads stranded (pass frac, blue), fulfills both criteria (pass both, green), or does not fulfill any of the criteria (none, gray) (C) Proportion of loci passing the three minimal locus profile rules (fn1, fn2, and fn3, described in table below). Arabidopsis thaliana (Artha) and Drosophila melanogaster (Drmel), low (LC) and high-confidence (HC) miRNAs from miRbase tested in a single library are included as a reference. (D) Number of loci without available source sRNA-seq libraries (light gray), loci failing one or more rules indicated in B and C (red), loci passing all rules (yellow), and loci passing all rules in two or more independent libraries (blue-green). For confirmed loci, we show in magenta those passing all the rules for annotating miRNAs in plants described in the ShortStack (Axtell, 2013b) and Axtell-2018 (Axtell and Meyers, 2018) rule sets. (E) Locus count and proportion of loci as in (D), separated by the parent species and including loci for which a valid hairpin could not be identified (dark gray).



