©1996-2019 All Rights Reserved. Online Journal of Bioinformatics. You may not store these pages in any form except for your own personal use. All other usage or distribution is illegal under international copyright treaties. Permission to use any of these pages in any other way besides the before mentioned must be gained in writing from the publisher. This article is exclusively copyrighted in its entirety to OJB publications. This article may be copied once but may not be reproduced or re-transmitted without the express permission of the editors.
Online Journal of Bioinformatics
Volume 5:13-22, 2004.
Prediction of Protein Homo-oligomers by Incorporating Quasi-Sequence-Order Effect.
Zhang SW, Pan Q, Zhang HC, Wu YH, Shi JU.
Department of Automatic Control,
Zhang SW Pan Q, Zhang HC, Wu YH, Shi J,U Prediction of Protein Homo-oligomers by Incorporating Quasi-Sequence-Order Effect, Onl J Bioinform., 5:13-22, 2004. A method of extracting features from protein primary sequences, incorporating quasi-sequence-order effects, was used to predict protein homodimers, homotrimers, homotetramers and homohexamers. It was found that overall predictive accuracy based on amino acid composition and sequence-order-coupling numbers with one-versus-rest method was 81.25%, or 3.89% higher than that based solely on amino acid composition under identical conditions. With the same attribute parameter set at CSO, composed of amino acids and sequence-order-coupling numbers, the overall predictive accuracy of all-versus-all method was 13.14% higher than that of one-versus-rest method. These results show that the method of extracting attribute vector from protein primary sequences by incorporating Quasi-Sequence-Order Effect is feasible and the primary sequences of homo -oligomers contain protein quaternary information. The results suggest that all-versus-all multi-class classification methods may be effective for predicting the higher homo-oligomers.
KEY WORDS: sequence-order-coupling numbers; all-versus-all; one-versus-rest; homo -oligomers