Anurekha G, Geetha P
Autism is an impairment of development in the central nervous system. It affects social coordination, emotions and motor activity of an individual. Many research works on autism detection have emerged in the recent times. The shortcomings of those works are, only limited amount of data is employed and an independent omics data is analyzed for prediction. Depth knowledge associated with the disorder has to be found for early detection. Multi-omics data analysis can provide more insight into the distinctiveness of the disorder. In the current existence of supervised machine learning algorithms have a greater notion in the field of medical sciences. In the proposed system, a novel technique which integrates genomics and proteomics data to classify the candidate gene from non-autistic gene using various machine learning algorithms Adaboost, Random Forest, J48 and KNN. The proposed work is tested on the standard dataset (NCBI and SFARI) and their performances are analyzed based on classification accuracy.
Autism, Impairment, Nervous System, Multi-Omics, Machine Learning.