Structural analysis of long single-stranded RNA molecules with atomic force microscopy imaging

Abstract

Characterization of the structure of long RNA molecules (>1 kb) is usually a time-consuming and tedious process. In this study, we have developed an imaging procedure for obtaining images of the extended secondary structures of long RNA molecules combined with automated MATLAB-based data processing algorithms for identification of the domain architecture of the molecules in these images. These algorithms include a molecule autoselection procedure based on height and area thresholding, a morphological thinning procedure to generate skeletons of the molecule in order to analyze the branched structure of the molecules, and a procedure to generate local volume profiles along the main chain of the molecule for identification of domains and prediction of the number of nucleotides comprising each domain. The single-molecule nature of this technique also allows for the identification of varying conformations of the molecule and assessment of the conformational flexibility of the identified domain organization.