Pred-O3, a web server to predict molecules, olfactory receptors and odor relationships

Area of Science:

• Olfactory neuroscience
• Computational chemistry
• Bioinformatics

Background:

• The sense of smell involves volatile molecules interacting with olfactory receptors, but the odorant-receptor-odor (O3) relationship is poorly understood due to complex combinatorial codes.
• A large family of olfactory receptors and numerous odorant molecules contribute to the complexity of olfactory perception.

Purpose of the Study:

• To design Pred-O3, a web server providing insights into olfaction by predicting olfactory receptors and odors associated with novel molecules.
• To develop AI-based predictive models for understanding the structure-odor-receptor relationships.
• To propose physicochemical features related to odors and olfactory receptors based on molecular structure encoding.

Main Methods:

• Development of predictive models using Artificial Intelligence (AI) on a dataset of 5802 odorant molecules and 7029 odor annotations.
• Utilizing structural models of 98 olfactory receptors for a systematic docking protocol to predict molecular binding.
• Encoding odorant molecule structures to identify relevant physicochemical features.

Main Results:

• AI models can suggest olfactory receptors and odors for new molecules.
• Physicochemical features related to odors and/or olfactory receptors are proposed based on molecular structure.
• The docking protocol predicts whether a molecule can bind to a specific olfactory receptor.

Conclusions:

• Pred-O3 offers a valuable tool for predicting odorant-receptor interactions and aiding in the design of new odorant molecules.
• The web server assists fragrance research and sensory neuroscience by providing insights into the complex mechanisms of olfaction.
• Pred-O3 is accessible online for researchers in olfaction and related fields.