Digital olfaction predicts mosquito behaviour

From our collaborative work with computational scientists at Google Brain, two exciting preprints have been released on BioRxiv. The first describes an accurate model for odor perception that holds true across the animal kingdom. Intriguingly, organisation of molecules in this odor map aligns with their organisation in metabolic pathways, suggesting odor perception is tuned to recognising the metabolic state of things. The second manuscript describes a deep learning and digital archaeology approach for mosquito repellent discovery. A large portion of the newly discovered repellents have activity against ticks as well.