Comparative genomic profiling for cellular miRNA loci in PEL. (A) Variation in miRNA gene copy number. Plotted is the range of dCT_U6 on the vertical and the 5% trimmed mean dCT_U6 on the horizontal axis. The 5% trimmed mean is the arithmetic mean excluding the top and/or bottom 5% of the data. Crosses indicate the values for each cellular miRNA. KSHV miRNAs are shown by open circles; EBV miRNA gene loci, by closed circles; and hsa-miR-34a, by the arrow. Range indicates the difference between the lowest and highest value for each miRNA across all PEL cell lines. If a particular miRNA gene is present in every single cell line at 2 copies per cell, the range is 0 or close to 0, since it reflects only the measurement error. A high range indicates that one or more of the cell lines have more than 2 copies per cell (amplification) or have no copy per cell (deletion). This indicates that one or more cell lines in the sample have sustained amplifications or deletions. These were identified by analyzing the individual scatterplots in Figure S1. (B) Representation of individual clusters of pre-miRNAs as obtained after unsupervised clustering. The axes represent the first 3 principal components of the dataset. The objective of principal component analysis is to test whether all the data are correlated, or if indeed significant patterns or groups of data exist and how many. It serves as a quality control tool for our analysis. As can be seen here, 4, no more and no less, well-separated clusters emerged. These are further analyzed in Figure 3. Individual pre-miRNAs are represented by dots. Cluster membership is indicated by lines and semitransparent hulls.