Hardik Shah

Publication Date

Fall 2012

Degree Type

Master's Project


Computer Science


The mechanism and role of RNA secondary structure elements in the replication and translation of human positive-strand RNA viruses remains poorly understood. These secondary structures are formed when a single RNA strand folds over and base pairs with itself, forming various types of loop structures. RNA strands fold into specific shapes. This unique shape for each nucleic acid chain is the most stable state it can adopt. The lower the energy, i.e., the fold with highest number of base pairs, the higher the stability of the structure. The Dynamic Programming technique, such as the one used in Nussinov- Jacobson algorithm, predicts the locations to fold the sequence to give us an optimal solution. But, the Nussinov algorithm does not necessarily generate the most stable structure and may produce scattered matches that are not biologically relevant. More complex algorithms are needed to solve this problem. Hence, we study Zuker’s energy minimization algorithm that uses thermodynamic details with dynamic programming principles at the core. Nussinov-Jacobson and Zuker algorithms give the maximum number of base pairs that the given RNA molecule might have upon folding onto itself. We analyze the outputs produced by both algorithms for small subsequences and compare the predicted structures. Using a sliding window approach, we focus on specific parts of RNA and analyze their structure. Studying the genomes of RNA viruses will give an insight into the nucleotide positions that determine the virulence of the different virus strains.