Figure 4.

Example and schematics of the driver and dimensions within BIP Index. Colors represent the four dimensions in the BIP index; abbreviations correspond to the drivers behind each dimension (see text). Each driver is a composite of a number of variables, or PCA scores for collinear variables, weighted by their MRA coefficients. Drivers composing each dimension are in turn weighted by a coefficient obtained from linear programming. The BIP Index for a country is the Euclidean distance from the origin along the axes defined by the dimensions (not shown. β, corrected correlation coefficient.

Ariño et al. BMC Bioinformatics 2011 12(Suppl 15):S4   doi:10.1186/1471-2105-12-S15-S4