I studied the evolution and cell biology of Paramecium tetraurelia—a model ciliate with over 40,000 distinct protein-coding genes resulting from as many as three ancient whole-genome duplication events. I was interested in the functional diversification of these gene duplicates at the level of protein localization, but the commonly used tools to study this were tedious. I instead applied a protein-correlation profiling approach to this system by way of generating a dozen sub-cellular fractions with different protein constituents due to the density of their resident organelle and then assayed these fractions using quantitative mass spectrometry. Each protein’s unique abundance profile provided evidence for its subcellular localization, and I used both supervised and unsupervised classification algorithms to cluster proteins together based on the similarity of these profiles to several hundred “marker proteins” which I manually curated. After expanding the protein inventory for numerous organelles by as many as a thousand proteins, I determined many features not previously understood or appreciated such as mosaic biochemical pathways, evidence for differential sorting mechanisms, and the abnormal evolutionary patterns of the mitochondrial proteome of ciliates. I developed a simple bioinformatic tool to probe spatial proteomics datasets more easily for proteins of interest. I demonstrate its applicability using a handful of well-characterized proteins in the budding yeast Saccharomyces cerevisiae as well as interesting proteins in less well-studied model systems like P. tetraurelia and the apicomplexan Toxoplasma gondii to both recapitulate known interactions and discover new ones. Finally, I look for large-scale evidence of gene duplicates relocalizing to new cellular compartments in P. tetraurelia and S. cerevisiae using this new dataset and a previously generated one, respectively. I find thousands of pairs of duplicates which are differentially identified and display evidence for subcellular divergence, and this seems to be largely decoupled from large changes in protein sequence but are instead associated with indels in their N-terminal peptide. These findings support the use of high-throughput proteomic techniques to determine evidence of functional divergence of gene duplicates. Taken together, this works provides a deep characterization of one of the largest unicellular proteomes in nature.