Artificial Evolution

Volume 3871 of the series Lecture Notes in Computer Science pp 247-258

Incorporating Knowledge of Secondary Structures in a L-System-Based Encoding for Protein Folding

  • Gabriela OchoaAffiliated withDepartment of Computer Science, Universidad Simon Bolivar
  • , Gabi EscuelaAffiliated withDepartment of Computer Science, Universidad Simon Bolivar
  • , Natalio KrasnogorAffiliated withSchool of Computer Science and I.T., University of Nottingham

* Final gross prices may vary according to local VAT.

Get Access


An encoding scheme for protein folding on lattice models, inspired by parametric L-systems, was proposed. The encoding incorporates problem domain knowledge in the form of predesigned production rules that capture commonly known secondary structures: α-helices and β-sheets. The ability of this encoding to capture protein native conformations was tested using an evolutionary algorithm as the inference procedure for discovering L-systems. Results confirmed the suitability of the proposed representation. It appears that the occurrence of motifs and sub-structures is an important component in protein folding, and these sub-structures may be captured by a grammar-based encoding. This line of research suggests novel and compact encoding schemes for protein folding that may have practical implications in solving meaningful problems in biotechnology such as structure prediction and protein folding.