Thermal shift assays (TSA) are commonly employed to screen a set of ligands, for example small molecule compounds, and find those that bind to a target protein. For this, a thermal unfolding experiment can be run on Prometheus. Samples containing just the protein, or the protein mixed with a ligand, are run and compared. A change in protein stability upon ligand addition indicates that the two molecules are likely binding. Typically, the inflection points (IP) or melting points (Tm) of the protein are shifted towards higher temperatures upon interaction with a ligand, giving the assay its name. The ratio value of the unfolded state usually remains similar, unless the ligand itself fluoresces. In this case, it is recommended to use the single wavelength data to evaluate binding. 

There is no general answer to the question whether an observed shift is significant. This depends mainly on the samples and their intrinsic variability. For well-behaved samples, even very small shifts (<1 °C) can be meaningful. It is recommended to run at least triplicates and base any decisions on statistics: a shift that is larger than 3x the standard deviation of the protein-only sample can be assumed to be significant.

A change in the particle's hydrodynamic radius measured with dynamic light scattering may further indicate the presence of an interaction. Furthermore, DLS data can help to identify aggregating compounds or compounds promoting protein aggregation.

While thermal shift assays are useful as a fast and convenient first screening tool to find promising candidates for molecular interactions, they are not suited to yield accurate information on binding affinities (the strengths of those interactions). For binding affinity determination, better suited methods like Temperature-Related Intensity Change/MicroScale Thermophoresis (TRIC/MST) and Spectral Shift Technology are recommended.