The Tm is the melting point of a protein, meaning it is the temperature at which 50% of the protein population is unfolded while the other half is still folded.

If a protein's stability changes, its Tm is likely to change, too. Tms therefore serve as a useful parameter for comparing protein stability across conditions, buffers, proteins and so on. Inflection points are used in a similar way. 

In the context of Prometheus measurements, inflection points and melting points (Tm) are almost the same, but not quite. Both serve as useful parameters for comparing protein stability across conditions, buffers, proteins and so on. Both terms are often used interchangeably by convention, and when comparing IPs and Tms derived from Prometheus data to other technologies, it becomes clear that in the vast majority of cases, the values are very similar or even identical. It is however important not to confuse the two. 

Inflection points are calculated relatively directly from the measured data, using only a smoothing algorithm and derivation as intermediate steps. These are relatively simple mathematical operations that can be performed on data without requiring any specific conditions to be met. They are also fast to perform, making them well-suited for automatic data analysis.

A Tm, on the other hand, can only correctly be derived from measured data via a thermodynamic fit. Fits can only be applied to data that fulfills certain requirements, and/or they require some guidance to work well. For example, a fit (as applied in PR.Panta Analysis) needs baselines. If an unfolding profile has unclear baselines, it will be very difficult to fit. Even for data where the baseline looks perfectly clear to the eye, it may be next to impossible to apply a fit automatically. In addition, fitting is mathematically complex and therefore takes longer. For these reasons, fitting of thermal unfolding data to determine Tm is not suited for automatic data analysis in PR.Panta Control. Instead, PR.Panta Analysis allows applying defined fit models within user-defined regions of interest, which provide the necessary guidance for the fit.

There is an additional well-known (but often neglected) caveat about the Tm: strictly speaking, a Tm can only be derived from equilibrium measurements since it is a thermodynamic parameter. Thermal unfolding experiments are often not in equilibrium, because irreversible aggregation processes keep removing unfolded protein from the reaction, making the unfolding process irreversible. However, thermal unfolding data is still widely used to derive IPs (and Tms, after applying a thermodynamic fit). Since such IPs (or Tms) are very similar or even identical to true Tms, all these terms and concepts are often used interchangeably by convention.

Another aspect that plays a role here is the heating rate used in thermal unfolding measurements. Most commonly, a heating rate of 1 °C/min is used, with Prometheus as well as other techniques. It is important to know that faster heating rates mean that the situation resembles equilibrium even less, which means that measured IPs (and Tms, after applying a thermodynamic fit) will be less similar to true equilibrium Tms.

For some other thermodynamic parameters derived from fits in PR.Panta Analysis, the same is true: they can technically only be derived from equilibrium measurements. But again, as with Tms, this is widely accepted, not just for Prometheus measurements but for many other technologies as well.