Abstract: Protein secondary structure prediction is one of the most important problems in bioinformatics. I}he protein secondary structure prediction accuracy plays an important role in the field of protein structure research. In this paper, using a Knowledge Discovery Theory based on the Inner Cognitive Mechanism (KDTICM) , an efficient protein seconda- ry structure prediction algorithm based on mixed-SVM ( support vector machine) approach was proposed. The algo- rithm makes full use of the evolutionary information contained in the physicochemical properties of each amino acid and a position- specific scoring matrix generated by a PSI-SEARCH multiple sequence alignment, secondary structure can be predicted at significantly increased accuracy. At last, the experiments were used to show the superior accuracy and gen- erality of the new algorithm than other classical algorithm.

Key words: Protein secondary structure prediction, Mixed-SVM method, Compound pyramid model