Synthetic genetic array (SGA) analysis coupled with high-content screening (HCS) in Saccharomyces cerevisiae has provided a wealth of information on functional genomics. Until recently, genetic interactions in SGA analysis have used colony size as a proxy for cellular fitness. Although this metric has proven to be robust, higher resolution phenotypes such as subcellular morphology cannot be assessed. Since there are various perturbations in S. cerevisiae in which mutant growth is normal despite morphological abnormalities within subcellular compartments, it would be of great benefit to the yeast community to complement existing colony size data with cell morphology data. Using the SGA-HCS approach, the Boone and Andrews Labs at the University of Toronto, have produced an image-based dataset of subcellular mutant phenotypes in the context of genome-wide perturbations. To analyze these massive datasets of 900,000 images, our labs have developed a machine learning strategy that has been able to successfully detect and classify about half of all the observed and published phenotypes. However, it is challenging to computationally analyze and model a total number of 100 classifiers for all the expected phenotypes with the existing pipeline. Thus, to complete the analysis, I expanded on optimizing the existing pipeline by constructing classifiers for missing phenotypes, and score the genes generating aberrant morphologies. So far, the optimized pipeline can classify a substantial amount of the mutant phenotypes as preliminary. In order to validate genes resulting with similar mutant phenotypes, I generated the morphology profiles for each gene and performed a preliminary trend analysis on morphology profiles by comparing them to the genetic interaction network to identify genes with high fitness but aberrant morphology. By obtaining complete profiles, we can construct a new informative network for our collections to use alongside the fitness scores from genetic interaction networks. Comparing the biological interpretations of the genetic interaction network and the morphology profiles can reveal further information on biological enrichment and functional analysis that might have been overlooked by the multiplicative model of the fitness measurements alone. This analysis will allow for the identification of connections between discrete biological processes, the prediction of novel gene function, and the generation of a clearer understanding of basic eukaryotic cell biology.