The dihedrals are similar to those in any peptide structure, and are defined in the same way. 

Article Solution structures of Fc epsilon RI alpha-chain mimics: A b...

Article Retroenantiomers— a close match indeed

 Thank you very much for your reply.

but i doubt that if i adjust dihedral angles the structure will change? in that way i can get a perfect circular ring?

Hi Nikhil

The backbone dihedral angles phi and psi are defined in a standard way as @David mentioned. Change them and you will change backbone conformation.

It sounds like you are building a cyclic peptide in silico. There are many combinations that will be "allowed" when forming a closed chain (fewer of course with very short chains). You may get a start by "stealing" backbone dihedrals from a cyclic peptide of identical length in the PDB. This will at least give you a closed backbone from which to model further. Otherwise, most structure modeling programs will allow you to build a closed structure.

The diehedrak angles are tne standard ones  PHI and PSI called Ramachandran angles.

They are used whatever the peptide strucre is. To serach for a perfect circle backbone structure is a non-sense and have no physical particular significance.

It is more interesting to look for specific gamma-turns or beta-turns in the case of molecular docking studies (compatibilities with the target)

The definition of dihedral angles for cyclic peptides is the same that is used for peptides and proteins, i.e., the 'phi' and 'psi' angles, as defined by Prof. G.N. Ramachandran. 'Phi' denotes the dihedral angle corresponding to the rotation about 'N - C(alpha) bond' of the peptide/protein backbone. 'Psi' denotes the dihedral angle corresponding to the rotation about 'C(alpha) - C bond' of the peptide/protein backbone. Cyclic peptides, especially the backbone cyclized peptides can be somewhat strained molecules. One of the criteria that can influence the extent of strain in the backbone cyclized peptides is, number of amino acid residues comprising the peptide. Smaller cyclic peptides composed of fewer amino acids than the bigger cyclic peptides can be more strained. The dihedral angles are indeed important to understand and visualize the 3-dimensional structure of the molecules. Variations in the dihedral angles (which are also called as 'torsional angles'), give rise to different conformations or conformers. And hence, changes in the dihedral angles can give rise to different 3-dimensional structures. If I remember correctly, there is a study published by Prof. C. Ramakrishnan (IISc, Bengaluru, India) on the conformation of cyclic peptide. If I get its citation, I shall share it here.