Abstract:

---

Traditional string pattern matching is well defined as finding similarities between two strings in either in compressed or uncompressed form. These techniques have good application in natural language processing, plagiarism detection, information retrieval etc. However, these algorithms cannot be directly applied to Biological string sequences without modification, this due to the inherent characteristics of these type of strings. For example, in RNA sequences, each string (nucleotide) is complementary to one another as such traditional matching algorithms will yield false results i.e. two physically similar sequences might not be structurally similar. To address this problem, this research proposes a Structural matching (s-matching) algorithm to achieve finding true similarity between RNA sequences. It is technique that finds true similarities between RNA string sequences taking into cognizant the complementary base pairing characteristics of the RNA nucleotides. Also, due to the exponential increased in the size of uncharacterized RNA sequences traditional matching algorithms yielded low optimal performances. To further enhance the optimality of the results, this research uses data compression techniques to compress these sequences before applying the structural matching algorithm. The proposed algorithms will help in finding the structure of the newly discovered RNA sequences that is critical for its functional inference with more optimal performances.

Keyward(s): Biological string sequences, inherent characteristics, S-mtching and RNA

Download