Secondary structure prediction

An important issue when trying to understand protein function is to know the actual structure of the protein. Many questions that are raised by molecular biologists are directly targeted at protein structure. The alpha-helix forms a coiled rod like structure whereas a beta-sheet show an extended sheet-like structure. Some proteins are almost devoid of alpha-helices such as chymotrypsin (PDB_ID: 1AB9) whereas others like myoglobin (PDB_ID: 101M) have a very high content of alpha-helices.

With CLC Genomics Workbench one can predict the secondary structure of proteins very fast. Predicted elements are alpha-helix, beta-sheet (same as beta-strand) and other regions.

Based on extracted protein sequences from the protein databank (http://www.rcsb.org/pdb/) a hidden Markov model (HMM) was trained and evaluated for performance. Machine learning methods have shown superior when it comes to prediction of secondary structure of proteins [Rost, 2001]. By far the most common structures are Alpha-helices and beta-sheets which can be predicted, and predicted structures are automatically added to the query as annotation which later can be edited.

In order to predict the secondary structure of proteins:

        Toolbox | Classical Sequence Analysis (Image gene_and_protein_analysis) | Protein Analysis (Image proteinanalyses)| Predict secondary structure (Image secondary_structure)

This opens the dialog displayed in figure 15.18:

Image secondary_structure_step1
Figure 15.18: Choosing one or more protein sequences for secondary structure prediction.

If a sequence was selected before choosing the Toolbox action, this sequence is now listed in the Selected Elements window of the dialog. Use the arrows to add or remove sequences or sequence lists from the selected elements.

You can perform the analysis on several protein sequences at a time. This will add annotations to all the sequences and open a view for each sequence.

Click Next if you wish to adjust how to handle the results. If not, click Finish.

After running the prediction as described above, the protein sequence will show predicted alpha-helices and beta-sheets as annotations on the original sequence (see figure 15.19).

Image secondary_structure_output
Figure 15.19: Alpha-helices and beta-strands shown as annotations on the sequence.

Each annotation will carry a tooltip note saying that the corresponding annotation is predicted with CLC Genomics Workbench. Additional notes can be added through the Edit Annotation (Image edit_annotation) right-click mouse menu. Removing annotations.

Undesired alpha-helices or beta-sheets can be removed through the Delete Annotation (Image delete_annotation) right-click mouse menu. Removing annotations.