TopoStats - A program for automated tracing of biomolecules from AFM images.

Joseph G Beton,Bart W Hoogenboom,Luzie Helfmann,Robert Gray,Maya Topf,Alice L B Pyne,Robert Moorehead,Agnel Praveen Joseph

doi:10.1016/j.ymeth.2021.01.008

Abstract

We present TopoStats, a Python toolkit for automated editing and analysis of Atomic Force Microscopy images. The program automates identification and tracing of individual molecules in circular and linear conformations without user input. TopoStats was able to identify and trace a range of molecules within AFM images, finding, on average, ~90% of all individual molecules and molecular assemblies within a wide field of view, and without the need for prior processing. DNA minicircles of varying size, DNA origami rings and pore forming proteins were identified and accurately traced with contour lengths of traces typically within 10nm of the predicted contour length. TopoStats was also able to reliably identify and trace linear and enclosed circular molecules within a mixed population. The program is freely available via GitHub (https://github.com/afm-spm/TopoStats) and is intended to be modified and adapted for use if required.

Highlights

The use of Atomic Force Microscopy (AFM) in structural biology has been increasing over the past 30 years; AFM is a versatile and accessible technique for directly imaging single biomolecules
We have demonstrated the power of TopoStats, our software package for automated AFM image correction, molecule identification and tracing
Using simple examples, such as DNA minicircles at a range of lengths, we have shown that TopoStats can identify and trace isolated molecules, providing precise measures of contour length

Summary

Introduction

The use of Atomic Force Microscopy (AFM) in structural biology has been increasing over the past 30 years; AFM is a versatile and accessible technique for directly imaging single biomolecules. The advances in the field were facilitated in large part by hardware development: A 100X increase in image acquisition times has allowed the visualisation of dynamic biological processes [1,2] This has been coupled with the development of more sensitive imaging modes and probes that can resolve the doublehelix of DNA [3] or the subunits of a macromolecular protein complex [4] using commercially available equipment. In addition to seeing these changes in molecular structure, direct imaging with the AFM facilitates the observation of rare molecular states and conformations within a snapshot of a heterogeneous population, for example visualising deviations in the DNA double-helix induced by supercoiling [7] These unique features of the AFM make it a versatile structural biology tool that can operate either standalone and/or complementing other techniques such as cryo EM and X-ray crystallography, where rare conformations of molecules are obscured by averaging

Objectives

Methods

Results

Conclusion