Proteins are inherently anisotropic macromolecules due to the variation in the surface-exposed amino acid side-chains on the protein surface. Protein phase diagrams describe the solution conditions where phase transitions, such as liquidsolid and liquid-liquid, occur. The position of these phase boundaries is dependent on both the intrinsic characteristics of the protein and on its environment. As a result, changes to the protein or its environment can significantly alter its phase diagram. Human γD-crystallin (HGD) is a protein found in the eye lens and is remarkably stable at high concentrations. HGD exhibits short-ranged attractive interactions, i.e. shorter than a quarter of the protein diameter. As a result, HGD undergoes liquidliquid phase separation (LLPS). The temperature (Tph) at which LLPS occurs can be used as a measure of the strength of attractive interactions between proteins in solution. The effects of chemically modifying specific amino acids (Lys-2 and Cys-110) on the surface of HGD with a (hydrophobic) small molecule fluorescent label were examined using both experiments and simulations. By measuring the LLPS temperature for modified proteins in mixtures with native protein, it was possible to determine how surface anisotropy and the chemical properties of the modifier changed the protein’s phase behaviour. Very low modified protein compositions (as low as xm = 0.0001) were sufficient to increase Tph significantly (~14 K), but both the position and type of fluorescent dye used influenced Tph. A numerical model was designed to explain the experimental observations and revealed that the increase in LLPS in the presence of modified protein was due to a new increase in attraction in the system. The effects of chemical modification were further examined by modifying HGD with PEGylated biotin at both the Cys and Lys positions on the protein surface. Even at high modified protein compositions (xm ~ 0.9), there was little change in Tph relative to unmodified HGD when the modification was performed at the Lys position. However, Tph increased significantly for thiol modified protein at a similar modified protein composition (~10 K). Neither modification had any impact on the structure of the protein relative to unmodified HGD. This study further highlighted how both the specific chemical modification and its position on the protein surface can change the impact to a phase boundary. Finally, the nucleation and growth of protein aggregates in solution and in cells was probed using double mutants of HGD. For solution based measurements of the P23VR58H mutant of HGD, it appears that aggregate growth produces monodisperse particle sizes, which appear to grow via monomer addition. When measurements are made in smaller volumes, surface effects lead to heterogeneous nucleation and polydisperse aggregates emerge. However, heterogeneous nucleation can be suppressed by surface treatments. Aggregate growth in cells was demonstrated for the P23TR36S mutant of HGD. Heterogeneous aggregates of GFP labelled protein were observed in cells for up to 4 days and these particles sizes were predominantly 1-2 µm in size. While some larger particle sizes were observed, these were very few in number. Significant cell death was associated with later particle growth stages, which warrants future investigation.