Adaptive weighted least squares method for the estimation of DNA fragment lengths from agarose gels

The size of DNA fragments is most frequently estimated from their electrophoretic mobilities. Agarose gels are used to estimate the size of DNA fragments ranging from a few hundred nucleotides to more than 20 kbp. The common practice when estimating the unknown fragment sizes is to plot the log of the size of molecular weight standards against their mobility and read the values of unknowns from this graph. However, due to perturbations in the gel, such plots often show pronounced curvature which may introduce significant subjectivity into the interpolation process. We present a new method "adaptive weighted least squares (AWLS)" based on the significance test to choose the order of polynomial. We compare this with the method introduced by Schaffer based on the modification of the Southern method. The results obtained by AWLS are significantly better than the method introduced by Schaffer. Different lanes are tested for consistency.