The Sound of Proteins
Sound and image are always found to be not only exciting, but also useful to characterise and distinguish objects. In particular, along with imaging, the sound rendering of various physiological signals (e.g. heart sound) has been seen as one of the diagnostic tools in medicine. A part from a few studie, in line with the advances in the post-genome era, the sonification concept at the molecular level (e.g., proteins) does not seem to have been taken into consideration.
In the recent years, signal processing techniques have been used in bioinformatics to extract information that is expected to reveal protein’s biological functions. Transforming proteins into signals, and then analyzing them with techniques such as the Discrete Fourier Transform (DFT) has revealed to be a particularly functional method for gathering information about the biological functions of those proteins. Specific similarities in the amplitude of particular frequency components have in fact been found within the signals generated from proteins having similar functions, being different from the ones generated by other proteins, reinforcing the idea that certain spectral components can be considered as a “signature” of particular protein’s families.
A further development of this technique is currently being undertaken, assessing the possibility of transforming those signals, whose lengths and frequencies are far out the audible range, in hearable sounds, with distinct spectral components. The advantages of such improvement could clearly be seen in the possibility given to the biologist to “listen” to certain proteins within a set, and to combine them, and/or set them apart, into different families considering their spectral components, hence the “timbre” of the different sounds.
Participants: Lorenzo Picinali, Huseyin Seker (Centre for Computational Intelligence, DMU), Charalambos Chrysostomou.