In this work we present and apply a 3-D pattern matching algorithm developed to solve structural recognition problems in molecular biology. The ability to automatically predict molecular interactions, is important in rational drug design and discovery, as well as a research tool in biomolecular structural recognition. Docking of a pair of molecules involves finding a proper partial fit between their molecular surfaces, as molecules interact at their surface. The generation of docked binding modes between two associating molecules depends on their 3-D structures and on their conformational flexibility, namely hinge-bending induced transitions. In hinge-bending, relatively rigid molecular subparts rotate on hinge(s) with respect to each other. This type of movements can occur in either of the two participating molecules, i.e., in the ligand or in the receptor molecule. Current automated docking techniques enable hinge bending flexibility in small ligands (e.g., drug molecules). Our approach allows hinge induced motions to exist in either diverse size ligands or variable size receptors (e.g., enzymes). We achieve this by the extension of the generalized Hough transform and Geometric Hashing techniques, which were originally developed for partially occluded articulated (flexible) object recognition in Computer Vision & Robotics. We show experimental results obtained by applying the algorithm on pairs of molecules allowing hinge-bending in either of the molecules for both complexed (bound) and unbound molecular configurations. We also discuss additional implications and applications of our approach.