Implementing receptive fields with excitatory and inhibitory optoelectrical responses of bacteriorhodopsin films

Abstract

The sign of the optoelectrical response of bacteriorhodopsin is highlighted as a means to emulate excitation and inhibition in neural computation. A classic example of a neural computation that is based on such excitation and inhibition is chosen to highlight the unique applicability of bacteriorhodopsin in highly parallel computational schemes. The classic example chosen is that of the ganglion receptive field, which is a fundamental element in retinal edge detection. Dried bacteriorhodopsin films are constructed that effectively act as receptive fields because of the sign of their photoresponse. The results on these simple bacteriorhodopsin receptive fields are extended to schemes that incorporate with greater elegance this unique ability of bacteriorhodopsin to exhibit excitation and inhibition. Experiments are presented that test some of these advanced ideas in bacteriorhodopsin parallel computation.