Abstract
The accurate and stable prediction of protein domains is an important stage for the prediction of protein structure, function, evolution, and design. Predicting inter-domain linkers is very important for accurate identification of structural domains within a protein sequence. This chapter provides a comprehensive comparative study of the existing approaches in domain linker prediction and discusses the technical challenges and unresolved issues in this field. It also provides a brief overview of protein structure, and discusses the evaluation measures that are typically used in domain linker prediction. The chapter then presents a comprehensive description and comparison for the most current research approaches in the field of inter-domain linker prediction. It outlines a domain boundary prediction using enhanced general regression network (DomNet) approach. Finally, the chapter reviews a domain linker prediction approach using amino acid (AA) compositional index and simulated annealing.
Original language | British English |
---|---|
Title of host publication | Pattern Recognition in Computational Molecular Biology |
Subtitle of host publication | Techniques and Approaches |
Publisher | wiley |
Pages | 212-235 |
Number of pages | 24 |
ISBN (Electronic) | 9781119078845 |
ISBN (Print) | 9781118893685 |
DOIs | |
State | Published - 28 Dec 2015 |
Keywords
- Compositional index
- Domain linker prediction assessment
- DomNet approach
- Inter-domain linker prediction
- Protein structure
- Simulated annealing