Numerical analysis of density gradient centrifugation profiles from eukaryotic DNA

A numerical method for the deconvolution of superimposed Gaussian distributions with a unique solution has been proposed by Medgyessy [10]. We have tested the usefulness of this method for the analysis of density gradient centrifugation profiles from eukaryotic DNA, which are normally composed from overlapping Gaussian distributed profiles of several subcomponents with different mean buoyant densities. From the analysis of human DNA and from model calculations we conclude that major subcomponents can be identified by this method, if they differ in their buoyant density by approximatly 0.005 g/ml. Minor components can only be identified if the total DNA has been fractionated according to buoyant density and the analysis is performed on the density profiles of the subfractions. This procedure represents a quick method to determine a reliable minimum number of subcomponents of DNA.