Tryptophan fluorescence wavelength is widely used as a tool to monitor chan
ges in proteins and to make inferences regarding local structure and dynami
cs. We have predicted the fluorescence wavelengths of 19 tryptophans in 16
proteins, starting with crystal structures and using a hybrid quantum mecha
nical-classical molecular dynamics method with the assumption that only ele
ctrostatic interactions of the tryptophan ring electron density with the su
rrounding protein and solvent affect the transition energy. With only one a
djustable parameter, the scaling of the quantum mechanical atomic charges a
s seen by the protein/solvent environment, the mean absolute deviation betw
een predicted and observed fluorescence maximum wavelength is 6 nm. The mod
eling of electrostatic interactions, including hydration, in proteins is vi
tal to understanding function and structure, and this study helps to assess
the effectiveness of current electrostatic models.