Various cold-adaptedorganisms produce antifreeze proteins (AFPs), which prevent the cell fluids from freezing.AFPs haveseveral important applications in increasing freeze tolerance of crop plants,maintain the tissue in frozen condition and producing cold-hardy plants using transgenictechnology. In this paper, we proposed a novel methodfor predicting AFPs usingSequential Minimal Optimization(SMO)classifier incorporation 4 types of features:hydropathy,physicochemical properties,amino acid composition and evolutionary profile. Testedby10-fold cross validation, our proposed method gains91.8accuracy. In addition, results reveal the better performance of our method in AFPs detection in comparison to the current state-of-the-art methods