We present an analysis of the blind predictions submitted to the fold recognition category for the second meeting on the Critical Assessment of techniques for protein Structure Prediction. Our method achieves fold recognition from predicted secondary structure sequences using hidden Markov models (H
Prediction of protein structure classes and secondary structures by means of hidden Markov models
โ Scribed by Hiroshi Yoshikawa; Mitsunori Ikeguchi; Shugo Nakamura; Kentaro Shimizu; Junta Doi
- Publisher
- John Wiley and Sons
- Year
- 1999
- Tongue
- English
- Weight
- 267 KB
- Volume
- 30
- Category
- Article
- ISSN
- 0882-1666
No coin nor oath required. For personal study only.
โฆ Synopsis
This study deals with structure class/secondary structure prediction of proteins using hidden Markov models (HMMs). With the proposed method, prediction is performed using HMMs designed so as to represent hierarchicality and periodicity of protein structural features. Secondary structures (partial tertiary structures) are formed from amino acid sequences while tertiary structures are formed through packing of secondary structures; thus, hierarchicality of protein structure is represented by means of hierarchical combination of multiple HMMs. In so doing, the tertiary HMM is built from a sequence of secondary structure segments while the secondary HMM is built from amino acid sequences. In addition, periodicity is introduced into the HMM network topology so that periodical structural features can be represented. Transition probabilities and output probabilities are determined through learning of data related to known structures. HMMs designed as mentioned were applied to the prediction of unknown structures. Accuracy above 50% was achieved for structure class prediction. Besides, high accuracy prediction of secondary structure was obtained for D class and D/E class. Thus, the proposed method proved to offer faithful representation of protein structural features.
๐ SIMILAR VOLUMES
A thermodynamic model describing formation of alpha-helices by peptides and proteins in the absence of specific tertiary interactions has been developed. The model combines free energy terms defining alpha-helix stability in aqueous solution and terms describing immersion of every helix or fragment
A new dataset of 396 protein domains is developed and used to evaluate the performance of the protein secondary structure prediction algorithms DSC, PHD, NNSSP, and PREDATOR. The maximum theoretical Q 3 accuracy for combination of these methods is shown to be 78%. A simple consensus prediction on th
Proteins of known structures are usually classified into four structural classes: all-โฃ, all-โค, โฃุโค, and โฃ/โค type of proteins. A number of methods to predicting the structural class of a protein based on its amino acid composition have been developed during the past few years. Recently, a componentc
Analysis of our fold recognition results in the 3rd Critical Assessment in Structure Prediction (CASP3) experiment, using the programs THREADER 2 and GenTHREADER, shows an encouraging level of overall success. Of the 23 submitted predictions, 20 targets showed no clear sequence similarity to protein