DNA Matching Using k - mer Derived Spatial Features

DNA sequence analysis has enormous applications including gene modification, gene therapy, new variety development, etc., due to which, the size of the genome datasets is increasing exponentially, and it is propagating more computational challenges. Existing DNA sequence analysis algorithms are two types, e.g., alignment-based (AB) and alignment-free (AF). AB is effective for short and homologous sequences, but time and memory complexity is extremely high where AF algorithms can solve the major limitations. Existing AF algorithms use different relative information, but these techniques lose spatial information. Therefore, this research proposes a novel AF algorithm by introducing two brand new features standard deviation and zero count which is efficient in terms of memory, time, and accuracy. At first, it generates a k - mer count matrix and a position vector for each cell of the count matrix. Later, it calculates the standard deviation of the first-order difference and the number of zero counts of the second-order difference of positions. The method is tested in several benchmark datasets and the performance is compared with existing studies and tools. For all datasets, it shows 1.217 to 354 times less memory complexity and top accuracy. In the case of time complexity, it is 20 to 5768 times faster than the MEGA tool. Therefore, this system can be an effective platform for DNA matching.