Can short time Fourier transform detect the localized latent periodicity of a protein sequence?

Spectral analysis has been used to detect latent periodicities of linear macromolecules, which are usually useful for the structural and functional study of those sequences. Those spectral analysis are largely based on Fourier transform which only can reveal long-range periodicities. In many cases, a set of arbitrarily chosen numbers are used to represent the constituting elements of DNA or protein in order to transfer a character sequence to a numerical series. Here, we introduce a time-frequency representation for protein sequences based on short-time Fourier transform (STFT) and electron-ion interaction potential (EIIP). We also demonstrate that the information related to a protein's biological properties for the chosen protein example could be revealed if the window length is chosen appropriately for the STFT.