Classify a Protein Domain using Sigmoid Support Vector Machine

Protein domains are portion block of protein sequence that evolved independent function. Therefore, the classification of protein domain is becoming very importance in order to produce new sequence with new function. However the main issue in protein domain classification is to classify the domain correctly into their category since the sequence coincidently classify to both category. Therefore, to overcome this issue, this paper proposed a computational method to classify protein domain from protein subsequences and protein structure information using sigmoid kernel function. The proposed method consists of three phases: pre-processing, protein structure information generating and post-processing. The pre-processing phase selects potential protein. The protein structure information generating phase used several calculations to generate protein structure information in order to optimize the domain signal information. The classification phase involves Sigmoid SVM and performance evaluation. The performance of the proposed method is evaluated in terms of sensitivity and specificity on single-domain and multiple-domain using dataset SCOP 1.75. This method showed an improvement of prediction in term of sensitivity, specificity and accuracy.