Adaptive optimizations performed using a genetic algorithm are employed to construct optimal laser pulse configurations that separate spectroscopic features associated with the two main energy-transfer pathways in the third-order nonlinear optical response simulated for the Fenna-Matthews-Olson (FMO) photosynthetic complex from the green sulfur bacterium Chlorobium tepidum. Superpositions of chirality-induced tensor components in both collinear and noncollinear pulse configurations are analyzed. The optimal signals obtained by manipulating the ratios of various 2D spectral peaks reveal detailed information about the excitation dynamics.