• Abstract
• Table of Contents
• Figures
• Literature Cited

Abstract

Recognition and prediction of structural domains in proteins is an important part of structure and function prediction. This unit lists the range of tools available for domain prediction, and describes sequence and structural analysis tools that complement domain prediction methods. Also detailed are the basic domain prediction steps, along with suggested strategies for different protein sequences and potential pitfalls in domain boundary prediction. The difficult problem of domain orientation prediction is also discussed. All the resources necessary for domain boundary prediction are accessible via publicly available Web servers and databases and do not require computational expertise. Curr. Protoc. Protein Sci. 66:2.14.1?2.14.16. © 2011 by John Wiley & Sons, Inc.

Keywords: structural domains; domain parsing; homology modeling; ab initio predictions; functional domains

        GO TO THE FULL PROTOCOL: PDF or HTML at Wiley Online Library Table of Contents

• Introduction
• What are Structural Domains?
• How Structural Domains are Defined
• Predicting Structural Domains
• Initial Steps in Identifying Protein Domains
• Methods for Domain Prediction
• Evaluating Domain Predictors
• Domain‐Domain Interactions
• Potential Problems
• Literature Cited
• Figures
• Tables

        GO TO THE FULL PROTOCOL: PDF or HTML at Wiley Online Library Materials

GO TO THE FULL PROTOCOL: PDF or HTML at Wiley Online Library Figures

• Figure 2.14.1 Domain parsing for two targets from the CASP structure prediction experiments. Target T0457 (A ) has two clearly defined domains without insertions with a short linker and few contacts between the domains. Target T0487 (the Argonaut silencing complex) is very complex to parse (B ). In the CASP8 experiment it was parsed into five domains (shown here in different shades); the domains were not linear in sequence and there were a number of domain extensions (or decorations). The orientation of the domains is complicated by the fact that the Argonaut silencing complex also binds DNA.

  View Image

• Figure 2.14.2 Four different ways to parse the same structure, target T0424 in CASP8. The beta‐sheet sub‐structure on the left in all the figures is a duplication, and although the duplication would be regarded as a single domain based on the strength of residue‐residue contacts between the sub‐units (B and C ), it could also be regarded as two domains from an evolutionary point of view (A and D ).

  View Image

• Figure 2.14.3 Flow chart summarizing the key steps involved in domain boundary prediction.

  View Image

• Figure 2.14.4 Two domains with a short linker where docking might be used with constraints to predict domain orientation. In fact, CASP target T0323 has an inserted domain (on the right) and therefore has two separate constraints that could have been used to limit the docking possibilities.

  View Image

• Figure 2.14.5 CASP target T0427, two domains joined by a long linker. The two domains are shown in different shades and the linker in black. Although the linking residues that join the two domains interact with the surface of domain 1 (shown on the left), there were no similar structures with these linking residues, so they could not have been modeled. The long linker even allows domain 2 to interact with the opposite face of domain 1, so here it would not have been possible to use docking constraints to limit the possibilities of interaction.

  View Image

• Figure 2.14.6 A difficult‐to‐predict case. CASP target T0547 has four domains shown in different shades. Although there are remotely similar structural templates for the two larger domains, the domain boundaries of the two smaller domains would have to have been predicted by ab initio methods.

  View Image

• Figure 2.14.7 Domain decorations. For each of these CASP targets there was a structurally similar template that could have been used for modeling for the regions shown in darker shades, but there were no templates for the regions shown in lighter shades. The regions without a template are domain decorations and are difficult to predict. Target T0510 (in A ) had two structural domains (shown in darker shades) and a small C‐terminal extension that more or less folded on its own. Target T0395 (B ) formed a single domain with a C‐terminal decoration that interacts with the domain surface and even forms a knot. Target T0409 (C ) is shown as a dimer. The N‐terminal extension interacts with the other chain in the dimer.

  View Image

Videos

Literature Cited