Abstract:
The field of neuroscience has made significant strides in understanding the mapping between physical properties of stimuli and perceptual characteristics in various sensory modalities such as vision and hearing. However, the olfactory system, responsible for our sense of smell, poses a unique challenge. Unlike the well-established mappings between wavelength and color or frequency and pitch, the relationship between chemical structures of molecules and their olfactory percepts, or odors, remains poorly understood. At the core of smell is our sense of olfactory information flow, in which smell is governed by the intricate relationship between odorant molecules and the subset of 400 olfactory receptors they activate. Combinations of unique receptor activations code for unique scents (combinatorial coding). Thus, understanding the intricate relationship between odorant molecules and their olfactory receptors (ORs) is crucial for unraveling the mysteries of human olfaction and its potential therapeutic applications. In this work, we investigate the use of geometric learning methods, both trained on proteins (olfactory receptors, specifically) and molecules, to enhance our understanding of odorant-receptor interactions and molecule-odor mappings, towards unveiling the combinatorial code of olfaction.
Bio:
Seyone Chithrananda is a 3rd year undergraduate student studying computer science at UC Berkeley. He is interested in using computation as a mechanism for understanding how biology works - and how we can engineer it to improve human health. His research goal is to build computational tools for the design, engineering & interpretation of biological systems. At Berkeley, Seyone conducts undergraduate research in Prof. Jennifer Doudna's laboratory, building machine learning methods for protein structural search and CRISPR engineering, and is a research intern at Microsoft Research working on problems relating to machine learning for proteins and molecules. Previously, he was a research intern at Dyno Therapeutics, working on machine learning for protein engineering, and worked in the laboratory of Prof. Alan Aspuru-Guzik at the University of Toronto on chemical language models.
29 сен 2023