We propose a pairwise protein structure alignment approach based on a joint similarity measure of multiple protein attributes. We map information on a protein's sequence location, structure and characteristic properties onto a highly-localized three-dimensional Gaussian waveform. By allowing the waveform to undergo unique transformations in the time-frequency plane, we allocate distinct parameters to represent the different attributes. Protein matching by expanding the mapped waveforms using appropriately designed basis waveform functions provides a similarity measure to encompass the multiple attributes. Simulations using data from a database demonstrate the performance of the joint alignment approach to infer relationships between proteins.