Hearing loss is one of the most prevalent human birth defects. Genetic factors contribute to the pathogenesis of deafness. It is estimated that one-third of deafness genes have already been identified. The current work is an attempt to find novel genes relevant to hearing loss using guilt-by-profiling and guilt-by-association bioinformatics analyses of approximately 80 known non-syndromic hereditary hearing loss (NSHL) genes. Among the 300 newly identified candidate deafness genes, slc26a2 were selected for functional studies in zebrafish. The slc26a2 gene was knocked down using an antisense morpholino (MO), and significant defects were observed in otolith patterns, semicircular canal morphology, and lateral neuromast distributions in morphants. Loss-of-function defects are caused primarily by apoptosis, and morphants are insensitive to sound stimulation and imbalanced swimming behaviours. Morphant defects were found to be partially rescued by co-injection of human SLC26A2 mRNA. All the results suggest that bioinformatics is capable of predicting new deafness genes and this showed slc26a2 is to be a critical otic gene whose dysfunction may induce hearing impairment.