Statistical feature selection from chaos game representation for promoter recognition

Authors:
Orawan Tinnungwattana;Chidchanok Lursinsap
Affiliations:
Advanced Virtual and Intelligent Computing (AVIC) Center, Department of Mathematics, Chulalongkorn University, Bangkok, Thailand;Advanced Virtual and Intelligent Computing (AVIC) Center, Department of Mathematics, Chulalongkorn University, Bangkok, Thailand
Venue:
ICCS'06 Proceedings of the 6th international conference on Computational Science - Volume Part II
Year:
2006

Citing 7
Cited 0

C4.5: programs for machine learning

C4.5: programs for machine learning
Induction of Decision Trees

Machine Learning
Characterization of Prokaryotic and Eukaryotic Promoters Using Hidden Markov Models

Proceedings of the Fourth International Conference on Intelligent Systems for Molecular Biology
Consistency-based search in feature selection

Artificial Intelligence
Promoter Recognition for E. coli DNA Segments by Independent Component Analysis

CSB '04 Proceedings of the 2004 IEEE Computational Systems Bioinformatics Conference
Integration of Knowledge-Discovery and Artificial-Intelligence Approaches for Promoter Recognition in DNA Sequences

ICITA '05 Proceedings of the Third International Conference on Information Technology and Applications (ICITA'05) Volume 2 - Volume 02
DNA sequence classification via an expectation maximizationalgorithm and neural networks: a case study

IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews

Quantified Score

Hi-index	0.00

Visualization

Abstract

The accuracy of promoter recognition depends upon not only the appropriate representation of the promoter sequence but also the essential features of the sequence. These two important issues are addressed in this paper. Firstly, a promoter sequence is captured in form of a Chaos Game Representation (CGR). Then, based on the concept of Mahalanobis distance, a new statistical feature extraction is introduced to select a set of the most significant pixels from the CGR. The recognition is performed by a supervised neural network. This proposed technique achieved 100% accuracy when it is tested with the E.coli promoter sequences using a leave-one-out method. Our approach also outperforms other techniques.