Pyrfume: A window to the world's olfactory data

Advances in theoretical understanding are often fueled by access to large, diverse experimental datasets. Olfactory neuroscience has lagged behind other sensory fields, largely due to the scarcity of comprehensive datasets that connect olfactory stimuli with corresponding percepts, behaviors, and neural circuits. To address this, we developed a unified framework, called Pyrfume, designed to integrate and standardize dozens of olfactory datasets, spanning different species and sensory modalities. This effort aims to provide a much-needed resource for researchers, accelerating progress in unraveling the complexities of olfaction and bringing the field closer to the level of understanding achieved in other sensory systems. The Pyrfume repository is notable for its breadth and coverage, spanning >40 odorant-linked datasets in mammalian olfaction including human psychophysics and perception; and animal psychophysics, behavior, brain imaging, physiology, and pharmacology. The foundation of this archive is built on two straightforward principles: 1) most olfactory experiments can be easily indexed by the unique identifier (ID) of an olfactory stimulus, whether it’s a molecule, substance, or mixture at a specified concentration; and 2) any olfactory experiment can be described in a universally readable format as a combination of chemical identifiers, the task performed with the stimuli, and the observed individuals along with their corresponding behaviors. By adhering to these principles, Pyrfume enables scientists, engineers, and trainees to delve into, evaluate, and explore the vast landscape of olfactory research and facilitates further cross-experiment analysis and discovery. This repository is also available to access via https://github.com/pyrfume.