The expected radius of gyration of the embedded coordinates can be calculated from the complete matrix of interatomic distances produced for the embedding. This expected radius is often larger or smaller than the radius of gyration calculated from the embedded coordinates. Therefore, the MMDG statement calculates a scaling factor equal to the ratio of the expected to actual radii of gyration and declares it to be the symbol $DGSCALE. The embedded coordinates should be multiplied by this factor, because it makes subsequent regularization easier.