Genomic Transformers: Innovative Approaches to Hepatitis Virus Subtyping

This article introduces a highly effective method for the classification of four subtypes of Hepatitis virus B, C, D, and E utilizing the sophisticated Transformer models. Initially, the DNA sequences are segmented into discrete non overlapping kmers, which effectively transforms the genetic information into a structured vocabulary. Subsequently, these segmented words are converted into vectorized representations through the innovative application of the Word2Vec technique. These detailed, highdimensional vectors capture the nuanced patterns within the sequences, which are then processed by powerful Transformer models for precise classification which results in classification accuracy of 99.49%. This advanced approach leverages the cutting-edge capabilities of Natural Language Processing to provide deep insights into the genetic basis of different Hepatitis virus' subtypes, marking a leap forward in bio-informatics.