On the performance of internal feedback artificial bee colony algorithm (IF-ABC) for protein secondary structure prediction

The prediction of protein secondary structures has been the focus of bioinformatics for decades. Since Christian Anfinsen´s pioneering work, the prediction scheme has been transformed into a numerical optimization case, which purses for the lowest free-energy value. Artificial bee colony algorithm (ABC) is a famous intelligence algorithm, and this work applies a novel ABC modified by internal feedback information (IF-ABC) for such optimization scheme. In IF-ABC, the convergence information is fully utilized as guidance for the subsequent convergence process. In addition, the original roulette selection strategy as in ABC is reasonably removed here. Comparable experiments are conducted on some sequences listed in the database of Protein Data Bank (PDB). Simulation results confirm, with statistical significance, that IF-ABC is more efficient and robust for such protein secondary structure prediction cases.