Classification of transmembrane segments in human proteins using wavelet-based energy

The special features of transmembrane segments (TMs) as well as the number of TMs within a protein sequence are often associated with properties related to the function or structure of the protein. A classification scheme for the TMs of human proteins is proposed in this work, by defining specific patterns of the total energy of each transmembrane segment after applying the continuous wavelet transform (CWT) to the hydrophobic sequence of the protein. The scheme was applied to proteins of known structure extracted by public available databases and statistical analysis followed in order to reveal possible patterns in the sequence of TMs within a protein. The results show some kind of selectivity in the occurrence of TMs with respect to the proposed types, thus allowing further analysis for the extraction of biological information of the proteins based on the proposed classification.