Next:
Introduction
Up:
No Title
Previous:
No Title
Contents
Contents
Introduction
Methods
The problem of objectively testing secondary structure prediction methods
Training and test sets of protein structures
Generating the multiple sequence alignments
Prediction methods analysed
Consensus Prediction Method
Assessment of accuracy
Results and Discussion
Comparison of secondary structure definition methods
Analysis of the test and training alignments
Alignment quality
Comparison of prediction accuracy for the CB396 test and RS126 training sets
Effect on
Q
3
of changing the number of related sequences
Effect on
Q
3
of reducing redundancy in multiple alignments
The effect on accuracy of alternative 8 to 3 state reductions
Single sequence prediction methods
Improving the consensus prediction
Summary and Conclusions
Acknowledgements
Bibliography
james@ebi.ac.uk