Accession Number:

ADA411639

Title:

A Hydrophobicity Based Neural Network Method for Predicting Transmembrane Segments in Protein Sequences

Descriptive Note:

Conference paper

Corporate Author:

SHANGHAI JIAO TONG UNIV (CHINA)

Personal Author(s):

• Chen, Zhongquiang
• Liu, Qi
• Zhu, Yisheng
• Li, Yixue
• Xu, Yuhong

Report Date:

2001-10-25

Pagination or Media Count:

5.0

Abstract:

Transmembrane proteins play vital roles in living cells. The difficulties in determining the topology of transmembrane protein experimentally and the increasing amino acid sequence data from genome projects provide great demand for computational methods to predict the region of transmembrane segments in protein sequences. A hydrophobicity based supervised learning vector quantization neural network prediction method is presented. The prediction accuracy is above 90 and comparable to existing methods.

Descriptors:

• *PROTEINS
• *TOPOLOGY
• NEURAL NETS
• PROBABILITY
• AMINO ACIDS
• CHINA
• HYDROPHOBIC PROPERTIES
• GENOME

Subject Categories:

• Biochemistry

Distribution Statement:

APPROVED FOR PUBLIC RELEASE