Prediction of Bio-informatics Data Related to Toxicant (Gallium Arsenide) using Artificial Neural Network

Toxic metals Gallium (Ga) and Arsenic (As) are widely found in our environment. Humans are exposed to these metals from numerous sources, including contaminated air, water, soil and food. Exposure to these semiconductors mostly occurs in microelectronic industry where the intermetallic compound Gallium Arsenide (GaAs) is frequently used during the preparation of material, cleaning and maintenance operations for quartz glassware. The toxic effect of this compound appears to occur due to inhalation or oral exposure and may result in poisoning. Recent studies indicate that it acts as a catalyst in the oxidative reactions of biological macromolecules and may lead to oxidative tissue damage. In this paper we present the results of an experiment conducted on rats to study the effect of the toxicant (GaAs) on the vital biochemical variables namely deltaaminolevulinic acid dehydratase (ALAD), blood glutathione (GSH/ GSSG) and reactive oxygen species (ROS). Subsequently an artificial neural network (ANN) based prediction model has been developed which is able to predict the changes in the biochemical variables of rats with high rate of accuracy. Hence it can effectively assist the scientists who perform biological experiments on animals for studying the effects of toxicity by replacing conventional process with an ANN based model. Keywords— Artificial Neural Network (ANN), Feedforward Back Propagation Algorithm, Bio-informatics Toxicant Data, Medical Expert System, MATLAB and toxicity prediction.