Table 2.

Fit parameters for anisotropy decay of protein-dye conjugates

System	τmol/ns	τfast/ns	Amol
Alexa488, experiment	3.80 (0.02)	0.24 (0.01)	0.52 (0.002)
Alexa488, ff03∗//TIP3P, N-terminal	1.31 (0.13)	0.06 (0.06)	0.78 (0.07)
Alexa488, ff03w//TIP4P/2005, N-terminal	5.03 (0.72)	0.47 (0.16)	0.59 (0.09)
Alexa488, ff03w//TIP4P/2005, C-terminal	3.48 (0.31)	0.11 (0.04)	0.81 (0.03)
Alexa488, ff03ws//TIP4P/2005(1.10), N-terminal	5.19 (1.73)	0.34 (0.21)	0.51 (0.10)
Alexa488, ff03ws//TIP4P/2005(1.10), C-terminal	3.70 (0.52)	0.18 (0.02)	0.56 (0.05)
Alexa594, experiment	3.55 (0.03)	0.46 (0.01)	0.43 (0.004)
Alexa594, ff03∗//TIP3P, N-terminal	1.57 (0.86)	0.34 (0.16)	0.57 (0.09)
Alexa594, ff03w//TIP4P/2005, N-terminal	3.29 (1.58)	0.47 (0.51)	0.73 (0.09)
Alexa594, ff03w//TIP4P/2005, C-terminal	7.79 (2.54)	0.76 (0.11)	0.29 (0.08)
Alexa594, ff03ws//TIP4P/2005(1.10), N-terminal	2.64 (0.26)	0.43 (0.05)	0.46 (0.06)
Alexa594, ff03ws//TIP4P/2005(1.10), C-terminal	3.65 (1.03)	0.50 (0.05)	0.41 (0.07)

Data were fit to the function $r\left(t\right)=r_0\left(\left(1-A_{\text{mol}}\right)e^{-t/{\tau }_{\text{fast}}}+A_{\text{mol}}\right)e^{-t/{\tau }_{\text{mol}}}$, where r₀ ≡ 0.38. The notation after the chromophore name is protein force-field//water model. Experimental data are for CspTm nonspecifically labeled with a single chromophore at either the N- or C-terminal labeling site. Note that in each simulation, the protein force field used was optimized to be used with the water model specified. We only list the water models explicitly for clarity. Errors for experiment are estimated by Monte Carlo resampling of the original data, while for simulation, errors were estimated from separate fits after dividing the data into 10 equal blocks.