Computational-Assisted and Data-Driven Delineation of Olfactory Cognitive Abilities

Context & Background

Olfaction (the sense of smell) is a critical human cognitive ability. Olfactory Receptors (ORs) are sensory receptors that detect odorant molecules. Unlike vision or hearing (which are measured in wavelengths), smell lacks a standardized metric, making it difficult to predict how molecules will smell or identify receptors involved in disease.

Problems to be Addressed

Over 400 human ORs are 'orphans,' meaning their target odorant molecules are unknown. Additionally, ORs are expressed in cancer cells (like prostate and gut tumors), where their role in tumor growth remains unmapped due to a lack of screening databases.

Aims and Objectives

1. Deorphanize olfactory receptors using Deep Learning.
2. Establish the EvOlf database of odor-receptor pairs.
3. Predict tumor-associated receptor activations using chemoinformatics.

Methodology

The project uses machine learning to classify chemical structures and screen them against OR sequences. Chemoinformatics tools (like Machine-Olf-Action) are upgraded to include explainable AI modules. Olfactory receptor-ligand bindings are validated experimentally using cell-line based assays.

Expected Outcomes

The EvOlf database, Chemoinformatics web dashboard for fragrance and healthcare industries, and tumor target candidates.