Dissertation > Excellent graduate degree dissertation topics show
Explorations of the Influence Factors on Accuracies of Protein Secondary Structure Prediction
Author: ZuoPengBo
Tutor: LiuJianGuo
School: Hebei University
Course: Biochemistry and Molecular Biology
Keywords: Protein Secondary Structure Prediction Hydrophobic factor HEC tendentious Double-layer SVM
CLC: Q51
Type: Master's thesis
Year: 2009
Downloads: 36
Quote: 1
Read: Download Dissertation
Abstract
|
From the corresponding three-dimensional structure of a sequence of the protein is one of the important issues in the field of bioinformatics . Computer prediction methods are widely applied to the study of protein secondary structure , the development process can be divided into two stages: the first stage of mathematical statistics as a starting point , based on the information of a single amino acid , such as Chou-Fasman and GOR ( Garnier - Osguthorpe - Robson) method ; second stage based on the evolutionary information, primarily in the sequence database search sequence using the BLAST tools such multiple alignment to obtain homologous information PSSM ( specific loci scoring matrix ) using the PSI-BLAST obtain the corresponding evolution information PSSM. The experimental committed to the characteristics of the amino acid -based PSSM prediction method improvements and improve prediction accuracy . SVM (Support Vector Machine) as a means of achieving , in the PSSM basis were added hydrophobic factor and HEC ( helical folding, random coil ) propensity two physicochemical factors as the characteristic values ??of the individual amino acids on the protein secondary structure prediction . This experiment is designed to double the SVM SVM use improved methods to achieve that through two methods of physical and chemical factors , and double- SVM tool to achieve the purpose to improve protein secondary structure prediction accuracy . The experimental results by the correlation coefficient analysis , the added hydrophobic factor and HEC tendentious Q3 weak positive correlation , a significant positive correlation with the SOV value . It proves that the hydrophobic amino acids HEC tendency to play a role in the formation of protein secondary structure . By double- SVM experiments , both the absolute accuracy or correlation coefficient analysis , double network advantage in the secondary structure prediction accuracy , the improved SVM their forecasting process play a significant role in optimization . The accuracy of the prediction value of Q3 and SOV than international PSSM method increased by 2.76% and 1.25% respectively .
|
Related Dissertations
- The Characteristic Analysis of the Protein Secondery Structure and Protein Interaction Prediction,Q51
- Purification and Stabilization of β-mannanase from Bacillus,TQ925
- Based on Genetic Algorithm for Protein Secondary Structure Prediction,Q51
- Computational Intelligence in Bioinformatics,TP399
- Artificial neural network methods in analytical chemistry,O651
- Research of Data Mining Technology Apply in the Protein Secondary Structure Prediction,TP311.13
- BA Scale-free Network Model’s Application and Extending,N941.4
- Biological Sequence Analysis Research Based on Data Mining,TP311.13
- Protein Secondary Structure Prediction Modeling Research Based on Neural Network,Q51
- The Research on the Spiking Neural Network and Its Applications,TP183
- Based of potato protein secondary structure prediction of neural network research,S532
- Protein Secondary Structure Prediction Methods,Q51
- Research of the Protein Secondary Structure Prediction Based on GA-BP Model,TP183
- Studies and Applications of Some Biological Sequence Analysis Algorithms,Q75
- A Study on the Protein Secondary Structure Prediction and the Connection between Protein Secondary Structure and Its 3D Structure,Q51
- Natural protein conformation and secondary structure analysis and forecast calculations,Q51
- Study on the Models of Protein Structure Prediction,Q51
- Prediction of Protein Secondary Structure and Interaction Based on Machine Learning,Q51-3
- Predicting Protein-Protein Interactions by Biostatistics Methods,Q51
- The Elective Toxic Effect of One Component Separated from Cobra Snake Venom on the Neuronal Cell and Its Possible Mechanisms,Q51
- Research on Predicting Intrinsic Disorder Protein Structure Based on Supervision Manifold Learning Algorithm,Q51
CLC: > Biological Sciences > Biochemistry > Protein
© 2012 www.DissertationTopic.Net Mobile
|