PepLook is a method designed to predict peptide 3D structures from amino acid sequences in different environments (1). PepLook is based on a Boltzmann stochastic approach exploring the conformational space of peptides by an iterative calculation that uses pairs of phi/psi angles determined by Etchebest et al. (2).The procedure generates 999 structures from which the model of lower energy is named 'the prime'. Using this method, we have suggested that cell penetrating peptides can adopt several conformations that help their penetration into the membrane (3).

Prediction for natural linear peptides:

Prediction for cyclic peptides:

Prediction for non-natural linear peptides:

(Predictions are in blue and pdb structures are in red)


1. Thomas, A., Deshayes, S., Decaffmeyer, M., Van Eyck, M. H., Charloteaux, B., and Brasseur, R. (2006) Proteins-Structure Function and Bioinformatics 65(4), 889-897

2. Etchebest, C., Benros, C., Hazout, S., and de Brevern, A. G. (2005) Proteins-Structure Function and Bioinformatics 59(4), 810-827

3. Deshayes, S., Decaffmeyer, M., Brasseur, R., and Thomas, A. (2008) Biochimica et biophysica acta 1778(5), 1197-1205

Username :
Password :

The project is financed by the Belgian Science Policy BelSPo under the IAP Phase VI Programme and is coordinated by the Centre de Biophysique Moleculaire Numerique of the University of Liege