iDEF-PseRAAC

Defensins are small cysteine-rich small cationic antibacterial proteins that act as non-specific mmune responses. The function of defensins helps to fight bacterial, viral and fungal infections, so this peptide is significantly increased in some pathogenic somatic cells. Because of their wide range of applications, predictive studies of the defensin families are very important.At the same time, the reduced amino acid is an effective method for optimizing sequence characteristics. In this work, we designed a predictor based on reduced amino acid composition called "iDEF-PseRAAC" to distinguish five types of defensin families. In the predictor, we use the reduced amino acid descriptors, 2-peptide composition, feature selection to optimize the sequence characteristics of the benchmark data, and the support vector machine is for the prediction process. Finally, the maximum accuracy we can get through 5 cross-validation is 90.48%. Hope the predictor will be helpful in defensin research.