The sensitivity, speed, and reproducibility of modern mass spectrometers enable in-depth new functional looks into the cellular proteome. Thousands of proteins can be detected in a single sample. In Co-Fractionation Mass Spectrometry (CF-MS) method, the input sample is fractionated by any biochemical method of choice. The reduced complexity of each fractionated sample leads to better proteome coverage. The separation profiles provide functional information on the proteins. This application has been used to predict organelle localization based on co-purification with marker proteins. More recently, CF-MS is being used to measure the apparent masses and determine the localization of soluble or membrane-associated protein complexes. This Ph.D. dissertation focuses on the extension of the boundary of CF-MS application to learn how protein complex evolution and protein complex composition have been accomplished. In the first part of this dissertation, the data will be presented on the degree to which variation in protein oligomerization across plant species is present, how proteomics in phylogenetic analysis (phyloproteomics/evolutionary proteomics) helps understand the evolutionary changes, and how oligomerization drives neofunctionalization during plant evolution. The latter part will describe that CF-MS coupled with multiple orthogonal chromatographic separations increases the resolving power of the profiling technique, enabling the composition of protein complexes to be predicted in the subaleurone layers of rice endosperm. Lots of novel protein complexes involved in RNA binding protein, translation, and the tissue-species metabolism will be discussed.