A new entropy model for RNA: part V, Incorporating the Flory-Huggins model in structure prediction and folding

Abstract

The effect of solvent-biopolymer interactions is hardly negligible. Whereas the ideal (non-interacting) polymer consisting of N monomers in an ideal solvent is expected to have the terminal ends of its chain with a root-mean-squared (rms) end-to-end separation distance (<em>rms</em>) proportional to the square root of <em>N</em>, real interactions of a <em>rms</em> polymer both with itself and with the solvent often tend to strongly perturb <em>rms</em>. In <em>rms</em> poor solvent, the biopolymer can collapse into a small globule much smaller than the ideal <em>rms</em> due to excluding solvent. In good solvent, the biopolymer can swell to a size much larger than the ideal r due to favoring solvent. These effects require rms corrections to an ideal polymer equation. We have been developing the cross linking entropy (CLE) model in this series. The model attempts find the maximum entropy of a folded polymer by taking into account the correlation caused by bonding and other interactions of the structure. In RNA, this mostly occurs in the stems. Here we adapt CLE model to handle polymer swelling and collapse for RNA molecules both in good and in poor solvent. This work is intended to introduce this type of study and to allow its systematic application in problems of RNA folding and structure prediction. The current study suggests that there may be some tendency for RNA to behave as a polymer in poor solvent and that this collapse may happen in sequences longer than 50 nt.