In this paper, a novel kernel taking into consideration of the physico-chemical properties of amino acids as well as the motif information is proposed to tackle the problem of protein classification. Similarity matrix is constructed based on an AAindex2 substitution matrix which measures the amino acid pair distance. Together with the motif content posing importance on the protein sequences, a new kernel is constructed. Numerical examples indicate that the string-based kernel in conjunction with SVM classifier performs significantly better than the traditional spectrum kernel method.
The folded k-spectrum kernel: A machine learning approach to detecting transcription factor binding sites with gapped nucleotide dependencies
