Over the years, we have developed several computational techniques for the statistical analysis of sets of electron micrographs of biological macromolecules. These methods include various types of factorial analysis (correspondence analysis, principal components), an outlier detection scheme, and clustering algorithms, as well as a statistical criterion for quantitative assessment of spatial resolution (spectral signal-to-noise ratio). An aspect that has been considered in more detail is the problem of correcting for magnification mismatches between micrographs due to varying imaging conditions. The motivation is to be able to obtain higher-resolution 3-D reconstructions of icosahedral viruses by combining a larger number of micrographs with slight disparities in magnification. For this purpose, we derived a fast iterative algorithm that determines the scaling factors by matching auxiliary one-dimensional functions (radial autocorrelation) computed for each micrograph. We have also developed a spline-based algorithm for scale conversion with an arbitrary scaling factor. The procedure is such that it minimizes the loss of information.
