Evolved Cellular Automata for Protein Secondary Structure Prediction

Phew…. Finally, the first draft of my research paper “Evolved Cellular Automata for Protein Secondary Structure Prediction” is complete. I guess the paper needs more revisions. But, I am feeling quite happy as this is my first research paper. Below is the abstract and full-text of the preprint.

Abstract: Cellular Automata has been used in this paper to predict the protein secondary structure. The research was inspired by the fact that cellular automata uses localized interactions to simulate global phenomena. The protein folding problem seems to be of exactly same nature; individual residues interact locally to give the protein chain a unique global conformation. The protein’s residue sequence was input to the cellular automaton and the rules for updating states were evolved using Genetic Algorithms to maximize Q3 on the dataset RS126. The maximum Q3 of around 58.2% was obtained, which although is low, but demonstrates the applicability of this simple technique on a problem as complex as protein structure prediction. There can be numerous possible improvements over this method, some of which are listed in the paper.

Click Here to read the complete paper in PDF format.

P.S: The program is coded in Python language. I will be releasing the code+dataset soon.

1 comment

  1. hi
    my name is javad mohammadzade
    I study Computer Science in University of tehran
    my new research is about Cellular automata and pridect protien and RNA folding
    when i search about in this field i visit your attempt “Evolved Cellular Automata for Protein Secondary Structure Prediction”
    I hope that help me and give me your research
    bye now

Comments are closed.