AudioGene (v4): Prediction of Deafness Genes by Machine Classification of Audiograms
Identifying genes that have similar effects on human disease is a key step in gene classification. In this study we consider the case of autosomal dominant non-syndromic hearing loss (ADNSHL) as a common human disease characteristic. AudioGene uses several well-known machine-learning classification methods to group audiograms (a common clinical measure of hearing loss) into groups that may correspond to a common gene mutation. While some of these machine learning methods are complementary, their use has parameter dependencies and relies on being presented with an unknown audiogram corresponding to one of the known genes. Our approach is to employ non-parametric methods to allow identification of “unknown” class results.
Concept of AudioGene:
