Characterisation of the noise affecting magnetic tweezers’ measures

Abstract

Magnetic tweezers are used to study and manipulate biomolecules by using a microscope and a CCD camera to track the position of micron-sized magnetic beads inside a microfluidic chamber. In these experiments we can attach the molecule of interest to a bead, and adjust the position of a pair of magnets placed above the chamber to apply different forces to the tethered molecule and track the molecule extension in real time. The resolution of these measurements are mainly affected by three sources of noise: the tracking noise, the Brownian motion noise and the low-frequency noise. Here we characterised these types of noise. We found that the tracking noise increases with the frequency at which the camera track as expected for a Poissonian uncorrelated noise. For the Brownian case, the noise increased with the length of the tethered molecule and decreases with the force applied. The dependencies found could be reproduced with a simple description of an inverted overdamped pendulum and using standard polymer models for the molecule elasticity. The low-frequency noise was present in all the experiments and could be corrected by substracting the average position of fixed beads inside the chamber