Melting Temperature Calculation

The melting temperatures of proteins are determined from the inflection points in the melting curve. In order to analyze complex unfolding profiles with multiple unfolding transitions, the PR.ThermControl uses smoothed differentiations of the data to precisely determine the inflection point for each transition in the melting curve (Figure 1).

For each curve, the second derivative is calculated, and its intercepts with the x-axis are determined. Relevant intercepts are determined based on the directionality of the melting curve. In the example depicted here, the melting curve is going upwards, and the second derivative is scanned for x-intercepts with the curve sloping downwards. This prevents the misinterpretation of irrelevant inflection points in the melting curve as melting points.

Figure 1: The left panel represents a simulated melting curve with two upward transitions in the fluorescence ratio. The framed part of the curve was differentiated (represented in the right panel). Here, the first and second derivatives of the curve are shown: f’ signifies the first and f’’ the second derivative (black and red curves, respectively). The PR.ThermControl software only shows the first derivative, but uses the second derivative to determine x-axis intercepts. Only the intercept points showing an opposite direction in the second derivative compared to the original curve are extracted as T_m.

Onset temperature determination

To determine the onset of aggregation or melting, the PR.ThermControl software uses an algorithm that looks for the very first two-state transition in the graph.

The onset corresponds to the temperature at which the two-state model fit deviates from the linear fit (baseline) by more than 0.5 % on this very first transition encountered.

Figure 2: Plot of the fluorescence ratio as a function of the temperature. The baseline linear fit (dotted black line) and the two-state model fit (black curve) extracted from the data set are shown, with the inflection point (T_m) represented as a cross. The 0.5 % deviation between those two fits is automatically calculated, and the corresponding temperature is plotted as the onset (red line, T_on).

This algorithm and the high density of data points collected by the Prometheus allow the PR.ThermControl to accurately determine the very first transition occurring in the sample and thus determining with precision the onset temperature for melting and/or aggregation.