A powerful technology has recently come into existence which will have profound implications for basic biology and medicine [16]. This technology is based upon the ability to “map” proteins on 2-dimensional electrophoretic polyacrylamide gels (2D gels), and for large-scale applications is highly dependent upon the use of computerized image analysis for the interpretation of the protein patterns displayed on the gels. For readers who may not be familiar with 2D gels it seems advisable first to describe briefly some of the essential facts about these gels. Basically, a 2D gel is made by distributing the proteins of a biological material in two dimensions, the vertical dimension reflecting the molecular weight of the protein and the horizontal dimension, its isoelectric point. Figure 1 is a photograph of a 2D gel prepared in the 2D gel lab at the University of Michigan and prepared from the solubilized protein contents of a sample of human red blood cells taken from a single individual. A given protein will migrate to characteristic, if somewhat approximate, x and y coordinates on the gel. In order to visualize where the proteins have migrated they are either labelled radioactively (followed by autoradiography) or stained with silver. In either case, the pattern of the migrated proteins can be digitized into computer readable format. The amount of stain or radioactivity is a function of the amount of protein. This produces a 3dimensional grey level surface where the third dimension indicates the amount of protein which migrated to the given planar coordinates. Thus, any one protein (characterized by a distribution around certain relative molecular weight and isoelectric point determined coordinates) will form a 3-dimensional peak or spot whose volume of integrated intensities is a measure of the amount of such protein which migrated to the given region. The digitized image of a 2D gel is a visual representation of a much larger number of proteins than has ever been visualized before. Thus, there are a large number of potential applications to both basic biology and medicine. In particular, a study now under way at the University of Michigan is attempting to estimate the rate with which human genes mutate [8]. Thousands of families, consisting of a father, mother, and a child, will contribute blood samples. Gels will then be made for each family member of each of various cell types within the blood, e.g., white and red blood cells. Each family trio of gels will then be digitized to produce computer-readable images of each gel in the family. A series of morphological algorithms [lo] are applied to the gel images in order to determine the locations of all the protein spots on each of the three gels. The three lists of protein spot