DNA-tract curvature profile reconstruction: A fragment flipping algorithm

Abstract

At nanometric level of resolution DNA molecules can be idealized with one dimensional curved line. The curvature value along this line is composed by static and dynamic contributions. The first ones constitute the intrinsic curvature, vectorial function of the sequence of DNA nucletides, while the second ones, caused by thermal energy, constitute the flexibility. The analysis of intrinsic curvature are a central focus in several biochemical DNA researches. Unfortunately observing this sequence-driven chain curvature, is a difficult task, because the shape of the molecule is largely affected by the thermal energy, i.e. the flexibility.A recent approach to this problem shows a possible methodology to map the intrinsic curvature along the DNA chain, observing an Atomic Force Microscopy image of a population of the DNA molecule under study. Reconstructing the intrinsic curvature profile needs a computing method to exclude the entropic contributions from the imaged profiles of molecules and to detect fragment orientation on image. The heuristic-search algorithm we propose can be a solution for these two tasks.