Feature reduction using a topic model for the prediction of type III secreted effectors

Authors:
Sihui Qi;Yang Yang;Anjun Song
Affiliations:
Department of Computer Science and Engineering, Information Engineering College, Shanghai Maritime University, Shanghai, China;Department of Computer Science and Engineering, Information Engineering College, Shanghai Maritime University, Shanghai, China;Department of Computer Science and Engineering, Information Engineering College, Shanghai Maritime University, Shanghai, China
Venue:
ICONIP'11 Proceedings of the 18th international conference on Neural Information Processing - Volume Part I
Year:
2011

Citing 3
Cited 0

Latent dirichlet allocation

The Journal of Machine Learning Research
High-accuracy prediction of bacterial type III secreted effectors based on position-specific amino acid composition profiles

Bioinformatics
LIBSVM: A library for support vector machines

ACM Transactions on Intelligent Systems and Technology (TIST)

Quantified Score

Hi-index	0.00

Visualization

Abstract

The type III secretion system (T3SS) is a specialized protein delivery system that plays a key role in pathogenic bacteria. Until now, the secretion mechanism has not been fully understood yet. Recently, a lot of emphasis has been put on identifying type III secreted effectors (T3SE) in order to uncover the signal and principle that guide the secretion process. However, the amino acid sequences of T3SEs have great sequence diversity through fast evolution and many T3SEs have no homolog in the public databases at all. Therefore, it is notoriously challenging to recognize T3SEs. In this paper, we use amino acid sequence features to predict T3SEs, and conduct feature reduction using a topic model. The experimental results on Pseudomonas syringae data set demonstrate that the proposed method can effectively reduce the features and improve the prediction accuracy at the same time.