SHAPE-RECOGNITION AND CHIRALITY MEASURE BY FITTING A DISCRETE PATTERNONTO A CONTINUOUS FUNCTION

This paper provides a new method for discrete shape similarity and dis similarity recognition. Instead of optimizing at once in discrete spac e, this method first optimizes the location of one of the discrete ent ities in a continuous approximation of the other. On the basis of the atom correspondences obtained, the two discrete shapes are then overla id. It is shown that the method can readily handle matching problems f or molecules with a few hundred atoms. It allows measurement of the ch irality of such systems as well as retrieval of a given included 3D pa ttern.