Biomolecular Recognition: Analysis of DFS Data - Dynamic Force Spectroscopy and Biomolecular Recognition - page 175

region (on the right of the curve, Figure 5.5) is set to 0. Furthermore, it is critical
to recalibrate the entire retract curve so that the slope of the hard contact zone (the
most left part of the curve, Figure 5.5) match the experimentally determined can-
tilever spring constant ( k cant ). This step is important for the appropriate calculation
of the effective loading rate (see Section 5.5).
The third step concerns the identification of rupture events. It varies from one
method to another. However, we have found it very difficult to analyze rupture events
directly from data points of the retract curve. The use of a derivative of the retract
curve (Sugisaki and Nakagini, 1999) or a sliding window approach was not con-
cluding in our hand to handle the large variety of observed retract curves at various
loading rates. Instead, a modified version of YieldFinder (Odorico et al. 2007b) con-
sists in modeling pulling events by a succession of small straight lines (Figure 5.6).
There are two main advantages: first, rupture events are defined by a change in the
sign of slope in two consecutive segments, and second, it makes it straightforward to
measure the loading slope before each rupture event.
The last step in FD curve analysis is the selection (or validation) of rupture
events and the building of an event database for further treatments (see below).
To select putative rupture events, several authors used different criteria such as

1.5

1.0

0.5

0

0 50 100 150 200 250 300 350 400 450

Z (nm)

FIGURE 5.5 Raw experimental retract curve obtained from a force-displacement curve. The
piezo displacement (Z) is shown in nanometer (nm). The Y-axis is a transformation of the
cantilever deflection into forces in n N. The tip displacement (piezo) of the retract curve starts
from the hard contact zone on the left, where the cantilever is bending due to the contact with
the hard substrate. Then, when the tip retracts, the cantilever stops bending when it crosses
the Y-axis at F = 0. Continuing the tip retraction shows a binding event represented by a
negative peak followed by an abrupt rupture and the return to the baseline toward the right
of the curve. To automatically detect rupture events, it is clear that a normalization of retract
curves is necessary.