Abstract
We have developed a new method for protein secondary structure prediction that achieves accuracies as high as 71.0%, the highest value yet reported. The main component of our method is a nearest-neighbor algorithm that uses a more sophisticated treatment of the feature space than standard nearest-neighbor methods. It calculates distance tables that allow it to produce real-valued distances between amino acid residues, and attaches weights to the instances to further modify the the structure of feature space. The algorithm, which is closely related to the memory-based reasoning method of Zhang et al., is simple and easy to train, and has also been applied with excellent results to the problem of identifying DNA promoter sequences.
Original language | English (US) |
---|---|
Pages (from-to) | 371-374 |
Number of pages | 4 |
Journal | Journal of molecular biology |
Volume | 227 |
Issue number | 2 |
DOIs | |
State | Published - Sep 20 1992 |
Keywords
- memory-based reasoning
- nearest-neighbor methods
- neural nets
- protein secondary structure
ASJC Scopus subject areas
- Biophysics
- Structural Biology
- Molecular Biology