We also noted that HL-1 cells expressed higher levels of miR-208a than miR-208b, seen as a marker of adult rather than embryonic cardiac tissue. There were also differences between the heart biopsy and HL-1 datasets, the most notable being the higher miR-145 levels and lower miR-1 levels in HL-1 cells. A recent study profiling SP600125 distributor abundant miRNAs in the whole mouse heart by llumina sequencing reported miR-1 as the most abundant miRNA, contributing,40% of tags. In the biopsy of the mouse left ventricle we also found that miR-1 was the most abundant miRNA, contributing 23% of tags. However in the atrial-derived HL-1 cell line miR-1 is the third most abundant miRNA contributing 6.3% of all miRNA reads. The 20 most abundant miRNAs from contributed 47.8% of miRNA reads for HL-1 dataset and 70.8% of the miRNA reads from our cardiac dataset, suggesting that the miRNA population of the whole heart is more similar to our ventrical biopsy than HL-1 cells. Furthermore, in adult tissue there is evidence that miR-145 is expressed to a much greater extent in the atrium versus the ventricle. As the ventricle contributes the bulk of material to the whole heart tissue as used in ref., and the entire material for our heart biopsy, we would suggest that the predominance of miR-145 in the HL-1 dataset is due to its atrial origin. In aggregate, these results indicate our detection of miRNAs in HL-1 cardiomyocytes is sensitive and consistent with expression in the adult heart. Our dataset confirmed the established view that most miRNA hairpins are asymmetrically processed, yielding mature miRNAs predominantly from one arm. Concordance with miRBase version 16 assignment of mature and miR* was generally good and we also found that -5p/-3p annotated miRNAs tended to be expressed relatively evenly. Nevertheless, expression was detectable at some level from both arms of most hairpins and, like others, we found that many individual hairpins markedly deviate in strand bias from their miRBase version 16 annotation. In some cases this appeared to reflect cardiomyocyteselective miRNA processing, in others we observed similar arm bias upon reanalysis of datasets from other tissues. Notably, the recently released miRBase version 18 has dispensed with the mature/miR* nomenclature and instead renamed all murine miRNA species with a -5p or -3p suffix, acknowledging the notion that expression from both arms can give rise to functional miRNA species. It is now appreciated that multiple miRNA features may affect miRNA strand selection. We found that no single aspect was uniquely required, however, two known features were sufficiently common to leave a ��signature�� in our analyses. First, using an unpaired 59 base as a surrogate measurement, we saw Bortezomib Proteasome inhibitor patterns consistent with the established thermodynamic stability rule for asymmetric miRNA incorporation into miRISC.