Towards Context-Aware DNA Sequence Compression Algorithms

High throughput DNA sequencing instruments significantly decreases sequencing costs and thus produces large volumes of genomic data. Thus for efficient storage and faster transfer of genomic datasets, compression is inevitable. DNA sequences are rarely accessed alone and DNA of same species have high degree of similarity. Thus considering each sequence separately and utilizing intra-similarity alone for compression fails to achieve impressive compression ratios when compared to direct encoding. Better compression ratios can be achieved if compression is done based on a reference sequence and encoding only the differences. This paper aims to overview various techniques developed so far for DNA sequence compression ranging from the initial general purpose compression algorithms to the latest multi-reference sequence based compression algorithms.