Algorithms and neural circuits in olfaction

Animals sense the chemical world to guide their behaviors. Fluctuating mixtures of odorants, often transported in fluid environments, are detected by an array of chemical sensors and parsed by neural circuits to recognize odor objects that can inform behavioral decisions. Unlike other sensory systems, the olfactory system lacks an obvious topographic organization, and neural connectivity across brain regions is seemingly unstructured. These anomalies offer an opportunity to uncover common principles across different sensory systems. We exploit a variety of biophysical, neurophysiological and behavioral methods to understand how odorant features are encoded in the activity of neurons and transformed in different stages of processing. We then use conceptual and computational models to seek algorithmic explanations for how animals solve specific olfactory tasks. In an illustrative set of studies, we have found that mice can be trained to recognize individual odorants embedded in unpredictable and variable background mixtures with high degree of success [doi: 10.1038/nn.3775]. Despite nonlinear interactions and variability in the representations of odor mixtures by odorant receptors, a simple linear feedforward decoding is sufficient to explain the performance of mice in this task [doi: 10.1016/j.neuron.2016.08.007]. Current experiments are aimed at understanding how the mouse brain represents information about odor mixtures to aid odor object identification and categorization.