A lot of practical problems can be formulated in terms of graph theory in as wide range of areas as theory of coding, data classification, computer vision and computational biology. Among them, the problem of finding maximum or maximal clique is often used. In this paper a modification of exact maximum clique algorithm is proposed and implemented. Results obtained on set of DIMAC benchmark problems are discussed. The results demonstrate the potential of modified algorithm to reduce computational burden when solving maximum clique problem. Main advantage of this modification is the possibility of usage as either heuristic or exact algorithm. Keywords: combinatorial problems, maximum clique algorithm, heuristic algorithm.
