Skip to main content
. 2014 Dec 30;9(12):e116221. doi: 10.1371/journal.pone.0116221

Figure 3. SVD trends in raw and normalized data.

Figure 3

SVD trends in raw (left panel) and normalized (right panel) clinical study data. Percentage at the top of each subplot shows the percent of the variation in the data explained by each trend. On the x-axis is sample index from 1 to 79, where each circle represents a sample. Samples are grouped by disease group with 39 control samples numbered 139 followed by 40 diabetes samples numbered 4079. Samples are in chronological run order on the instrument within each group. Values on the y-axis satisfy equation: R = UDV’ where R is a matrix of residuals (left panel) and matrix of normalized intensities (right panel); columns of V represent trends observed in the data. Three trends that explain the most amount of variation are plotted for raw and normalized data.