TY - JFULL AU - Yi-Zhong Weng and Chien-Kang Huang and Yu-Feng Huang and Chi-Yuan Yu and Darby Tien-Hao Chang PY - 2009/6/ TI - Introducing Sequence-Order Constraint into Prediction of Protein Binding Sites with Automatically Extracted Templates T2 - International Journal of Computer and Information Engineering SP - 1455 EP - 1462 VL - 3 SN - 1307-6892 UR - https://publications.waset.org/pdf/2272 PU - World Academy of Science, Engineering and Technology NX - Open Science Index 29, 2009 N2 - Search for a tertiary substructure that geometrically matches the 3D pattern of the binding site of a well-studied protein provides a solution to predict protein functions. In our previous work, a web server has been built to predict protein-ligand binding sites based on automatically extracted templates. However, a drawback of such templates is that the web server was prone to resulting in many false positive matches. In this study, we present a sequence-order constraint to reduce the false positive matches of using automatically extracted templates to predict protein-ligand binding sites. The binding site predictor comprises i) an automatically constructed template library and ii) a local structure alignment algorithm for querying the library. The sequence-order constraint is employed to identify the inconsistency between the local regions of the query protein and the templates. Experimental results reveal that the sequence-order constraint can largely reduce the false positive matches and is effective for template-based binding site prediction. ER -