Using Multilayer Perceptron Computation to Discover Ideal Insect Olfactory Receptor Combinations in the Mosquito and Fruit Fly for an Efficient Electronic Nose

Abstract

The model organism, Drosophila melanogaster, and the mosquito Anopheles gambiae use 60 and 79 odorant receptors, respectively, to sense their olfactory world. However, a commercial "electronic nose" in the form of an insect olfactory biosensor demands very low numbers of receptors at its front end of detection due to the difficulties of receptor/sensor integration and functionalization. In this letter, we demonstrate how computation via artificial neural networks (ANNs), in the form of multilayer perceptrons (MLPs), can be successfully incorporated as the signal processing back end of the biosensor to drastically reduce the number of receptors to three while still retaining 100% performance of odorant detection to that of a full complement of receptors. In addition, we provide a detailed performance comparison between D. melanogaster and A. gambiae odorant receptors and demonstrate that A. gambiae receptors provide superior olfaction detection performance over D. melanogaster for very low receptor numbers. The results from this study present the possibility of using the computation of MLPs to discover ideal biological olfactory receptors for an olfactory biosensor device to provide maximum classification performance of unknown odorants.

Improving odorant chemical class prediction with multi-layer perceptrons using temporal odorant spike responses from drosophila melanogaster olfactory receptor neurons

“…We have successfully used the Artificial Neural Network (ANN) approach in our previous work for predicting descriptor values of unknown chemicals [6], identifying chemical classes of unknown odorants [7][8][9][10][11] and predicting odorants and fruit blends of varying concentrations [5]. Thus far, we have used the spontaneous scalar data of receptor responses to odorants [4].…”

“…A mutant olfactory receptor neuron (ORN) was used to house the Ors and record their responses to the odorants. We grouped the 34 odorants into 4 different groups defined by their chemical class (with the number of chemicals listed in parentheses): Lactone (5); Aldehyde (4); Ester (14) and Alcohol (11). Prior to the MLP operation, both temporal and spontaneous data were preprocessed by zeromean and normalization to enhance network learning [12].…”

“…This scalar value representing DmOr responses was used to represent receptor characteristics [4,13,14] and used for analysis in our previous work [5,7,[9][10][11]15]. The spontaneous data contains the same 34 chemical odorants, 4 chemical classes and 6 DmOrs used in the temporal data.…”

“…In this work, we employed a feed forward Multi-Layer Perceptron (MLP) architecture [5][6][7][8][9][10][11] consisting of a double hidden layer hybrid MLP system (Fig. 4).…”

“…By combining MLPs into a hybrid system, we can address the multiclass problem associated with odorants of different chemical class [18]. From our previous work [5,8,10,11], we have found that a hybrid MLP architecture consisting of single-output double hidden layer MLPs to outperform single hidden layer MLPs with multiple output neurons. …”