Limits...
Using neural networks and evolutionary information in decoy discrimination for protein tertiary structure prediction.

Tan CW, Jones DT - BMC Bioinformatics (2008)

Bottom Line: Various features are extracted from the training dataset of positive and negative examples and used as inputs to the neural networks.This neural network is the best among all methods tested in discriminating the native structure from a set of decoys for all decoy datasets tested.This method is demonstrated to be viable, and furthermore evolutionary information is successfully used in the neural networks to improve decoy discrimination.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Computer Science, University College London, London, UK. chingwai@ntu.edu.sg

ABSTRACT

Background: We present a novel method of protein fold decoy discrimination using machine learning, more specifically using neural networks. Here, decoy discrimination is represented as a machine learning problem, where neural networks are used to learn the native-like features of protein structures using a set of positive and negative training examples. A set of native protein structures provides the positive training examples, while negative training examples are simulated decoy structures obtained by reversing the sequences of native structures. Various features are extracted from the training dataset of positive and negative examples and used as inputs to the neural networks.

Results: Results have shown that the best performing neural network is the one that uses input information comprising of PSI-BLAST 1 profiles of residue pairs, pairwise distance and the relative solvent accessibilities of the residues. This neural network is the best among all methods tested in discriminating the native structure from a set of decoys for all decoy datasets tested.

Conclusion: This method is demonstrated to be viable, and furthermore evolutionary information is successfully used in the neural networks to improve decoy discrimination.

Show MeSH

Related in: MedlinePlus

Enrichment scores (15% × 15%) produced by the S combination of the NN-solvpairndist, NN-solvpair, NN-dist methods, the K Nearest Neighbours methods (K = 10, K = 100) and the pairwise potentials method on the different individual decoy datasets, including the combination of all the individual datasets.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2267779&req=5

Figure 4: Enrichment scores (15% × 15%) produced by the S combination of the NN-solvpairndist, NN-solvpair, NN-dist methods, the K Nearest Neighbours methods (K = 10, K = 100) and the pairwise potentials method on the different individual decoy datasets, including the combination of all the individual datasets.

Mentions: Figure 4 shows the enrichment scores [16] of the S combination across all decoy datasets for the different methods. For the combined datasets, the pairwise potentials method [17] has the highest enrichment score [16], while the NN-solvpairndist method is comparable to the rest of the other methods. For most of the decoy datasets, there is no clear outstanding method which produces a distinctly high enrichment score [16], apart from the pairwise potentials method [17] in the Baker [16], 4state_reduced [24] and fisa_casp3 [4] datasets.


Using neural networks and evolutionary information in decoy discrimination for protein tertiary structure prediction.

Tan CW, Jones DT - BMC Bioinformatics (2008)

Enrichment scores (15% × 15%) produced by the S combination of the NN-solvpairndist, NN-solvpair, NN-dist methods, the K Nearest Neighbours methods (K = 10, K = 100) and the pairwise potentials method on the different individual decoy datasets, including the combination of all the individual datasets.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2267779&req=5

Figure 4: Enrichment scores (15% × 15%) produced by the S combination of the NN-solvpairndist, NN-solvpair, NN-dist methods, the K Nearest Neighbours methods (K = 10, K = 100) and the pairwise potentials method on the different individual decoy datasets, including the combination of all the individual datasets.
Mentions: Figure 4 shows the enrichment scores [16] of the S combination across all decoy datasets for the different methods. For the combined datasets, the pairwise potentials method [17] has the highest enrichment score [16], while the NN-solvpairndist method is comparable to the rest of the other methods. For most of the decoy datasets, there is no clear outstanding method which produces a distinctly high enrichment score [16], apart from the pairwise potentials method [17] in the Baker [16], 4state_reduced [24] and fisa_casp3 [4] datasets.

Bottom Line: Various features are extracted from the training dataset of positive and negative examples and used as inputs to the neural networks.This neural network is the best among all methods tested in discriminating the native structure from a set of decoys for all decoy datasets tested.This method is demonstrated to be viable, and furthermore evolutionary information is successfully used in the neural networks to improve decoy discrimination.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Computer Science, University College London, London, UK. chingwai@ntu.edu.sg

ABSTRACT

Background: We present a novel method of protein fold decoy discrimination using machine learning, more specifically using neural networks. Here, decoy discrimination is represented as a machine learning problem, where neural networks are used to learn the native-like features of protein structures using a set of positive and negative training examples. A set of native protein structures provides the positive training examples, while negative training examples are simulated decoy structures obtained by reversing the sequences of native structures. Various features are extracted from the training dataset of positive and negative examples and used as inputs to the neural networks.

Results: Results have shown that the best performing neural network is the one that uses input information comprising of PSI-BLAST 1 profiles of residue pairs, pairwise distance and the relative solvent accessibilities of the residues. This neural network is the best among all methods tested in discriminating the native structure from a set of decoys for all decoy datasets tested.

Conclusion: This method is demonstrated to be viable, and furthermore evolutionary information is successfully used in the neural networks to improve decoy discrimination.

Show MeSH
Related in: MedlinePlus