Interestingly, adipose core miRNAs that organized in clusters were both from the same (e.g. cluster including let-7a, -7d, -7f in chr. 8) or different miRNA families (e.g. cluster including miR-17-3p, -18a, -19a, -20a, -19b, -92 in chr. 12).
The combined analysis using both Pearson correlation coefficients with WGCNA and hierarchical clustering with Euclidean distance resulted in clusters that are both similar in transcript abundance and level among genotypes across time points.
This result was expected, as all the sequences in these clusters are similar in both composition and length, making them easy to align.
Although many of the genes in Cluster 1 are annotated as cell wall proteins (Supplementary Material), and although the two genes in Cluster 6 are both cyclins, the genes within these clusters have different expression patterns to one another (Fig. 4c, panels 1 and 6).
First, while the three QTL are presented as if they were the only traits in each cluster, in fact, both clusters are in genomic regions already known to be crowded with QTL.
Logically, for the newly formed cluster, it assigns a value of "1" with another cluster only when all the units in both clusters are connected.
Homotypic clusters are prevalent in both bacterial and eukaryotic genomes, and are present in both promoters as well as more distal enhancer/silencer elements.
Moreover, Sox9 staining patterns were mainly nuclear in normal cartilage, but Sox9 staining in OA clusters was present in both the cytoplasm and nucleus, or even exclusively in the cytoplasm.
The DSB clusters are found in both coding regions (i.e., genes) and in intergenic regions (IGRs), but they localize preferentially to large IGRs.
This suggested that members of these clusters are astrocytic in both NT2-A and primary human cells.
