Fig 7.tif (1.51 MB)
Download file

Principal components of NMJ morphometry.

Download (1.51 MB)
posted on 2016-03-21, 05:28 authored by Bonnie Nijhof, Anna Castells-Nobau, Louis Wolf, Jolanda M. Scheffer-de Gooyert, Ignacio Monedero, Laura Torroja, Lluis Coromina, Jeroen A. W. M. van der Laak, Annette Schenck

The correlation-matrix represents the pair wise correlations between all possible feature pairs, displayed as scatter plots (bottom left side) or as correlation coefficients (top right side) (A). The matrix is ordered so that those with the highest correlation are closer together. Correlation strength is color coded, ranging from light reddish (no-weak correlation) to an intense red color (for strongly correlating features). Five principle components were determined and the contribution of each feature to a certain component is summarized, together with the percentage of variance that is explained by that principal component and the cumulative percentage of total variance (B). The most contributing features per principal component (PC) are indicated in red. The first two components, which together make up for 59.83% of total variance are plotted with the first principal component (PC1) on the x-axis and the second principal component (PC2) on the y-axis (C). The arrow length reflects the contribution of each feature to the first two principal components: features showing a major contribution to PC1 or PC2 are indicated in black, and NMJ features with a minor contribution are grey.