In this work, we present a computational and statistical study of transcriptomic data, in the context of the model organism Saccharomyces cerevisiae. This study is related to recent advances in the understanding of how genes interact in a cell to produce complex phenotypes, where we start to give higher importance to the Gene Regulatory Network as a whole, instead of attributing phenotypic variation to mutations in a small group of genes. Hopefully, our work brings a fresh perspective, combining optimization, probabilistic graphical models, and complex networks, with the objective of showcasing the promise of this shift in understanding, by partially supporting the Omnigenic Hypothesis.