Breathing dynamics based parameter sensitivity analysis of hetero-polymeric DNA

Abstract

We study the parameter sensitivity of hetero-polymeric DNA within the purview of DNA breathing dynamics. The degree of correlation between the mean bubble size and the model parameters are estimated for this purpose for three different DNA sequences. The analysis leads us to a better understanding of the sequence dependent nature of the breathing dynamics of hetero-polymeric DNA. Out of the fourteen model parameters for DNA stability in the statistical Poland-Scheraga approach, the hydrogen bond interaction $\epsilon_{hb}(\mathtt{AT})$ for an $\mathtt{AT}$ base pair and the ring factor $\xi$ turn out to be the most sensitive parameters. In addition, the stacking interaction $\epsilon_{st}(\mathtt{TA}-\mathtt{TA})$ for an $\mathtt{TA}-\mathtt{TA}$ nearest neighbor pair of base-pairs is found to be the most sensitive one among all stacking interactions. Moreover, we also establish that the nature of stacking interaction has a deciding effect on the DNA breathing dynamics, not the number of times a particular stacking interaction appears in a sequence. We show that the sensitivity analysis can be used as an effective measure to guide a stochastic optimization technique to find the kinetic rate constants related to the dynamics as opposed to the case where the rate constants are measured using the conventional unbiased way of optimization.