Fig. 7: COLIS results from protein nucleation studies.

a Measured temperature as a function of time. The results of five independent runs are shown. The horizontal line at the bottom is the target final temperature. The inset shows that the undershoot is minimal. b Normalized absorbance from five repetitions of the cycle of cooling to 10.5 °C followed by heating to 40 °C. Dashed lines are fits of each curve to a sigmoid (only using data up to the maximum of each curve). c Experimental and fitted DLS correlation functions measured using DLS90 during a period of 30 s, at the end of the melting cycle at 40 °C. d distribution of cluster sizes obtained from the CONTIN fit shown in (c). The smaller peak is centered at 2.4 nm which is typical of a lysozyme molecule. The second peak, at 125 nm, is typical of the protein-rich clusters often observed prior to crystal nucleation.