Biology Reference
In-Depth Information
1000
Diffraction
pattern used
for 3-D
tracking
100
10
Bead
DNA
1
0
4
8
12
Separation distance (mm)
FIGURE 6.5
Calibration for high-force magnetic tweezers device. The main figure shows the calibration
curve of force versus the separation distance between the magnet assembly and the
sample plane. Data represented by circles are obtained by measuring the velocity at which
4.5-
m diameter beads moved through a glycerol solution of known viscosity under force.
Data represented by triangles are obtained by measuring the stretching of a single DNA strand
of known length while force is applied to the tethered magnetic bead (schematically depicted
in lower inset). Upper inset shows typical diffraction pattern for magnetic bead.
Adapted from Lin and Valentine (2012a) . Reprinted with permission from the
American Institute of Physics, Copyright 2012.
m
6.1.3.3 Particle tracking
Accurate tracking of bead position in three dimensions is essential both for proper
calibrations and for analysis of data of beads moving through the MT networks. This
tracking is achieved using Fast Fourier Transform-based particle tracking algorithms
implemented in LABVIEW (National Instruments), as previously described ( Gosse &
Croquette, 2002; Ribeck & Saleh, 2008 ). An autocorrelation of each Fourier image is
used to determine the xy centroid position of the bead. Axial position detection is
achieved using semicoherent, parallel illumination to generate a diffraction ring pat-
tern that depends sensitively on the distance of the particle from the focal plane. Each
image is compared to a lookup table of previously acquired calibration images, and
interpolation is used to find the best z coordinate. Because gel microrheology measure-
ments typically involve large particle displacements, this calibration image stack
typically consists of
m
m with
10 reference images collected per micrometer in the image plane. We have found that
using a computer with a fast, large cache memory is particularly important for the rapid
processing of the large image stacks required for 3D tracking of beads. The intrinsic
accuracy of the 3D tracking algorithms for beads immobilized in a low-contrast media
(such as water) is
300 images, where the total calibrated distance is
30
1 nm. When embedded in MT gels, additional noise sources are
Search WWH ::




Custom Search