Browsing by Author "Barsainyan, A.A."

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    Deep Learning for Odor Prediction on Aroma-Chemical Blends
    (American Chemical Society, 2025) Sisson, L.; Barsainyan, A.A.; Sharma, M.; Kumar, R.
    The application of deep-learning techniques to aroma chemicals has resulted in models that surpass those of human experts in predicting olfactory qualities. However, public research in this field has been limited to predicting the qualities of individual molecules, whereas in industry, perfumers and food scientists are often more concerned with blends of multiple molecules. In this paper, we apply both established and novel approaches to a data set we compiled, which consists of labeled pairs of molecules. We present graph neural network models that accurately predict the olfactory qualities emerging from blends of aroma chemicals along with an analysis of how variations in model architecture can significantly impact predictive performance. © 2025 The Authors. Published by American Chemical Society.
  • No Thumbnail Available
    Item
    Dense Sense: a novel approach utilizing electron density augmented machine learning paradigm to understand the complex odour landscape
    (Royal Society of Chemistry, 2025) Saha, P.; Sharma, M.; Balaji, S.; Barsainyan, A.A.; Kumar, R.; Steuber, V.; Schmuker, M.
    Olfaction is a complex process where multiple nasal receptors interact to detect specific odorant molecules. Elucidating structure–activity-relationships for odorants and their receptors remains difficult since crystallization of the odor receptors is an extremely difficult process. Therefore, ligand-based approaches that leverage machine learning remain the state of the art for predicting odorant properties for molecules, such as the graph neural network approach used by Lee et al. In this paper we explored how information from quantum mechanics (QM) could synergistically improve the results obtained with the graph neural network. Our findings underscore the possibility of this methodology in predicting odor perception directly from QM data, offering a novel approach in the machine learning space to understand olfaction. This journal is © The Royal Society of Chemistry, 2025