Protein Structure Prediction – Primary and Secondary Structure Analysis and Prediction

There are various tools for predicting the physical properties using the sequence information. Some of the major ones are discussed below:

1. Compute pI/Mw.

2. ProtParam.

3. SAPS.

4. Predicting protein hydrophobicity.

5. PEST and PESTfind.

6. Coils, Paircoil and Multicoil.

1. Compute pI/Mw:

This tool calculates the isoelectric point and molecular weight of an input sequence.

2. ProtParam:

It is a tool which allows the computation of various physical and chemical parameters for a given protein.

3. Predicting protein hydrophobicity:

ProtScale can be used to predict the hydrophobicity of a protein.

4. PEST and PESTfind:

PEST stands for Proline, Glutamic acid, Serine and Threonine residues.

PEST identifies possible PEST regions in a submitted probe using molecular fraction of the P, E, S and T components and the hydrophobicity index of the region.

5. Coils, Paracoils and Multicoil:

This program compares the sequence of a database of known parallel two-stranded coiled-coils and derives a similarity score.

Paircoil predicts the location of coiled-coil regions in the sequence.

Mulitcoil program predicts the location of coiled-coil regions in the amino-acid sequences and classifies the predictions as dimeric or trimeric.

Secondary structure analysis and prediction

The important secondary structure prediction methods are:

1. Chou-Fasman method:

This depends on assigning a set of prediction values to a residue and then applying an algorithm to the conformational parameters and positional frequencies.

2. GOR (Garnier, Osguthorpe and Robson) method:

The accuracy of prediction by using this method is around 64%. This method works better for helix than for sheet.

3. Nearest neighbor method:

This method is based on the hypothesis that short homologous sequences of amino-acids have the same secondary structure tendencies.

4. Hidden Markov Models:

This is used to predict the secondary structure of a protein of a given structural class e.g. alpha+beta as used in structural classification databases.

5. Multiple alignments based self-optimization method:

This is a secondary structure prediction program that uses multiple alignments.

2-D structure prediction

This is done in two steps:

1. Predicting inter-residue contacts:

Inter residue contacts can be predicted based on hydrogen bonding pattern between residues. Analyses of correlated mutations are done to predict long-range inter-residue contacts.

2. Predicting inter-strand contacts:

Prediction of inter-residue contacts can be simplified by predicting the contacts between residues in adjacent strands. This can be done based on potentials of mean-force.


Protein structure prediction which is one of the very useful and important application in bioinformatics can be done if the amino acid sequence of the protein is known. A number of protein identification and characterization tools are available. However, predicting the structure and functions of transmembrane helices, a special class of proteins that include GPCRs, is much needed, as they are important for therapeutic interactions. Although excellent tools and computational methods are available, none of the techniques os full proof and the area remains a very exciting one for the researchers.