DocumentCode :
3048517
Title :
A Novel Approach Predicting the Signal Peptides and Their Cleavage Sites
Author :
Yao, Lixiu ; Xue, Li ; Liu, Hui ; Chou, Kuo-Chen
Author_Institution :
Inst. of image Process. & Pattern Recognition, Shanghai Jiaotong Univ., Shanghai
fYear :
2007
fDate :
6-8 July 2007
Firstpage :
390
Lastpage :
393
Abstract :
The sliding window method will cause the severe unbalanced dataset problem. In this paper, under-sample the majority class method is adopted to solve this problem, and SVM is used to classify the processed data. The better prediction result of minority class (that is, the signal peptides positive sample set) is observed. Besides, we discover that the (-3,-1) rule is helpful to the prediction. So Information content based feature weighting method is proposed. This method avoids the blindness of the previous algorithm in dealing with different sites. Experiments show that not only is the correct prediction rate of minority class improved dramatically, but also the correct prediction rate of majority class is kept in a high level. Combination of the unbalanced data processing and the proposed information content based feature weighting method can greatly improve the performance of SVM classifier of signal peptides.
Keywords :
biology computing; molecular biophysics; pattern classification; proteins; support vector machines; SVM; cleavage sites; information content based feature weighting; protein signal; signal peptide classification; sliding window method; unbalanced data processing; Amino acids; Data processing; Image processing; Pattern recognition; Peptides; Proteins; Sequences; Signal processing; Support vector machine classification; Support vector machines;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Bioinformatics and Biomedical Engineering, 2007. ICBBE 2007. The 1st International Conference on
Conference_Location :
Wuhan
Print_ISBN :
1-4244-1120-3
Type :
conf
DOI :
10.1109/ICBBE.2007.103
Filename :
4272587
Link To Document :
بازگشت