X. Wang, J. T. L. Wang, D. Shasha, B. Shapiro, S. Dikshitulu, I. Rigoutsos, and K. Zhang. Automated discovery of active motifs in three dimensional molecules. In ########### ## ### ### ############# ########## ## ######### ######### ### #### ######, 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....as well by [2] 3] 14] Application of data mining to discover frequent substructure patterns in three dimensional graphs is not novel; however, previous research has primarily targeted small molecules and has not addressed issues of scalability and minimization of noise effects. The work of [19] detects substructures of three dimensional graphs, but the algorithm does not consider atom type, which due to steric and electrostatic behavior is critical to the quality of discovered substructures. 5] more explicitly targets the chemical domain by considering atom and bond types as well as ....

X. Wang and et. al. Automated discovery of active motifs in three dimensional molecules. In KDD, 1997.

....ii) e#cient algorithms that implement these techniques, and iii) the applications of above techniques in business and marketing domains. More recently, researchers have started tackling the problem of mining scientific data. In particular approaches for mining astronomical [5, 16] biological [8, 21, 41], chemical [7] and fluid dynamical [11] datasets have been recently proposed by various researchers. Few of the above methods actually account for the kinematically and dynamically driven characteristics of the data. Abstract representations in the form of graphs are often extracted and variants ....

....interactions among the coherent structures across time. However, such an abstract approach cannot exploit many of the inherent physical and dynamical properties of flow fields. For MD simulation data, we initially experimented with variants of the molecular substructure discovery algorithms [7, 8, 41]. We adapted the technique by Parthasarathy and Coatney [26] for mining biological macromolecules in protein data, in order to locate defects in silicon bulk lattices. The technique relies on range pruning and candidate pruning for reducing the search space of possible frequent substructures. ....

X. Wang, J. T.-L. Wang, D. Shasha, B. A. Shapiro, S. Dikshitulu, I. Rigoutsos, and K. Zhang. Automated Discovery of Active Motifs in Three Dimensional Molecules. In 3rd ACM SIGKDD Intl. Conference on Knowledge Discovery and Data Mining, pages 89--95, August 1997.

....when the clusters are of different tightness, and (ii) the degree to which they can lead to reasonably balanced clusters. I Introduction The topic of clustering has been extensively studied in many scientific disciplines and over the years a variety of different algorithms have been developed [31, 22, 6, 27, 20, 35, 2, 48, 13, 43, 14, 15, 24]. Two recent surveys on This work was supported by NSF CCR 9972519, EIA 9986042, ACI 9982274, by Army Research Office contract DA DAAG55 98 1 0441, by the DOE ASCI program, and by Army High Performance Computing Research Center contract number DAAH04 95 C 0008. Related papers are available via ....

Xiong Wang, Jason T. L. Wang, Dennis Shasha, Bruce Shapiro, Sitaram Dikshitulu, Isidore Rigoutsos, and Kaizhong Zhang. Automated discovery of active motifs in three dimensional molecules. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining, pages 8995, 1997.

....when the clusters are of different tightness, and (ii) the degree to which they can lead to reasonably balanced clusters. 1 Introduction The topic of clustering has been extensively studied in many scientific disciplines and over the years a variety of different algorithms have been developed [31, 22, 6, 27, 20, 35, 2, 48, 13, 43, 14, 15, 24]. Two recent surveys on # This work was supported by NSF CCR 9972519, EIA 9986042, ACI 9982274, by Army Research Office contract DA DAAG55 98 1 0441, by the DOE ASCI program, and by Army High Performance Computing Research Center contract number DAAH04 95 C 0008. Related papers are available via ....

Xiong Wang, Jason T. L. Wang, Dennis Shasha, Bruce Shapiro, Sitaram Dikshitulu, Isidore Rigoutsos, and Kaizhong Zhang. Automated discovery of active motifs in three dimensional molecules. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining, pages 89--95, 1997.

....in automatic partitioning of the meshes to cut out some parts that represent repeating patterns. Our technique uses the strategy of growing patterns bottom up from vertices. Bioinformatics community has been interested in automatic discovery of repeating structural motifs in molecular structures [21, 16]. However, they limit the discovery to a fixed class of structures. To enable effective classification of vertices in a given model, we associate a footprint 1 with each vertex in the model. To form a footprint we must select invariant properties #### over the features # # # in the model such ....

Xiong Wang, Jason Wang, Dennis Shasha, Bruce Shapiro, Sitaram Dikshitulu, Isidore, and Kaizhong Zhang. Automated discovery of active motifs in three dimensional molecules. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining (KDD97), California, pages 89--95, August 1997.

No context found.

X. Wang, J. T. L. Wang, D. Shasha, B. A. Shapiro, S. Dikshitulu, I. Rigoutsos, and K. Zhang. Automated discovery of active motifs in three dimensional molecules. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining, pages 89--95, Newport Beach, California, 1997.

....are classified according to their functions. However, recently, many approaches have been proposed to classify proteins according to their structures, e.g. sequences [6] secondary structures [6] and three dimensional structures [9] Many of these methods complemented the traditional approach. In [8, 9], we developed an algorithm for discovering frequently occurring patterns in three dimensional data and applied it to protein classification. While we succeeded in classifying two families of proteins with high recall and precision, experimental results showed that it was di#cult to extend the ....

....and their local coordinate systems SF are recovered based on the global coordinate system that defines the candidate pattern. The triplet matches are augmented to larger substructure matches when their recovered local coordinate systems match each other. The interested readers are referred to [8, 9] for more details. n1 n2 n3 p SF x y z y x z n5 n4 Figure 4. A substructure. n3 SF y x z n5 n4 o Figure 5. One of the triplets. For each family i of the proteins, we identify two types of patterns on the surfaces of the training data, the pro patterns and the con patterns. ....

X. Wang, J. T. L. Wang, D. Shasha, B. A. Shapiro, S. Dikshitulu, I. Rigoutsos, and K. Zhang. "Automated discovery of active motifs in three dimensional molecules," Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining, pages 89--95, Newport Beach, California, 1997.

....approximately occurring, within one mutation, in all the three graphs. Our strategy to find the patterns in a set of 3D graphs is to decompose the graphs into rigid substructures and then use geometric hashing [14] to organize the substructures and then to find the frequently occurring ones. In [35], we applied the approach to the discovery of patterns in chemical compounds under a restricted set of edit operations including node insert and node delete, and tested the quality of the patterns by using them to classify the compounds. Here, we extend the work in [35] by (i) considering more ....

....occurring ones. In [35] we applied the approach to the discovery of patterns in chemical compounds under a restricted set of edit operations including node insert and node delete, and tested the quality of the patterns by using them to classify the compounds. Here, we extend the work in [35] by (i) considering more general edit operations including node insert, delete and relabeling; ii) presenting the theoretical foundation and evaluating the performance and efficiency of our pattern finding algorithm; iii) applying the discovered patterns to classifying 3D proteins, which are ....

[Article contains additional citation context not shown here]

X. Wang , J. T. L. Wang, D. Shasha, B. A. Shapiro, S. Dikshitulu, I. Rigoutsos, and K. Zhang, "Automated Discovery of Active Motifs in Three Dimensional Molecules," Proc. 3rd Int'l Conf. Knowledge Discovery and Data Mining, Newport Beach, Calif., 1997, pp. 89--95.

....distance measures on trees and graphs and approximate query processing algorithms to support inexact matching. Develop a framework for selectivity estimation for queries on trees and graphs with wildcards. Develop a framework for turning searching to pattern discovery in trees and graphs [33, 94, 95, 100]. Develop support for semantic extensions: semi exible or exible queries [56] in which parent child relationships in queries may become ancestor descendant or even descendant ancestor relationships in data graphs. 5. ACKNOWLEDGMENTS We thank Ken Abe, Greg Heil, Katherine Herbert, Huiyuan ....

X. Wang, J. T. L. Wang, D. Shasha, B. A. Shapiro, S. Dikshitulu, I. Rigoutsos, and K. Zhang. Automated discovery of active motifs in three dimensional molecules. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining, pages 89-95, 1997.

....with deep implications [1, 2, 3] Traditionally, proteins are classified according to their functions. However, recently, many approaches have been proposed to classify proteins according to their structures, e.g. sequences [3] secondary structures [3] and three dimensional structures [6] In [5, 6], we developed an algorithm that discovers frequently occurring patterns in a set of 3D graphs. We applied the algorithm to protein classification. While we succeeded in classifying two families of proteins with high recall and precision, experimental results showed that it was difficult to extend ....

.... Family # obtains a score # # # # # #### # # ### # # # # ### ### # # 1 A candidate pattern # matches a substructure # with # mutations if by applying an arbitrary number of rotations and translations as well as # node insert delete operations to # , one can transform # to # (see [5, 6] for details) where # # # # # if # # occurs in # # otherwise The protein # is classified to the family # with maximum # . We add the denominator to make the score fair to all families. Notice that the maximum possible score for any family is 1. This is necessary because the three families ....

X. Wang, J. T. L. Wang, D. Shasha, B. A. Shapiro, S. Dikshitulu, I. Rigoutsos, and K. Zhang. Automated discovery of active motifs in three dimensional molecules. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining, pages 89--95, Newport Beach, CA, 1997.

No context found.

Wang, X.; Wang, J. T. L.; Shasha, D.; Shapiro, B. A.; Dikshitulu, S.; Rigoutsos, I.; Zhang, K. Automated discovery of active motifs in threedimensional molecules. Proceedings of the 3rd International Conference on Knowledge DiscoVery and Data Mining, 1997;pp89-95.

No context found.

X. Wang, J. T. L. Wang, D. Shasha, B. Shapiro, S. Dikshitulu, I. Rigoutsos, and K. Zhang. Automated discovery of active motifs in three dimensional molecules. In ########### ## ### ### ############# ########## ## ######### ######### ### #### ######, 1997.

No context found.

Xiong Wang, Jason T. L. Wang, Dennis Shasha, Bruce Shapiro, Sitaram Dikshitulu, Isidore Rigoutsos, and Kaizhong Zhang. Automated discovery of active motifs in three dimensional molecules. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining, pages 89--95, 1997.

No context found.

Xiong Wang, Jason T. L. Wang, Dennis Shasha, Bruce Shapiro, Sitaram Dikshitulu, Isidore Rigoutsos, and Kaizhong Zhang. Automated discovery of active motifs in three dimensional molecules. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining, pages 89--95, 1997.

No context found.

X. Wang, et al. Automated discovery of active motifs in three dimensional molecules. In KDD97.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC