Nucl2Vec : Local alignment of DNA sequences using Distributed Vector Representation

The Next Generation Sequencing Technique (NGS) has provided affordable and fast method for generating genetic data. Generation of whole Genome Sequence and extract relevant information from this data is still a computationally expensive process. In this paper we demonstrate a novel approach for local alignment of DNA reads with respect to reference genome. For this process we have used Skip-gram model for creating encoding(Nucl2Vec) and k-nearest neighbor for the alignment. With our new approach we have reduced computation cost for local alignment, while achieving accuracy comparable to existing defacto standard BWA-MEM tool.Index TermsGenome, Alignment, Local Alignment, k-nearest Neighbor, Distributed vector representation, Skip-gram