This thesis describes a nuclear shell model code which aims for a significant reduction in computer resource usage while retaining accuracy of results as compared to numerically exact solutions. I begin with an introduction to the configuration interaction and shell model calculations. I then motivate the need for a proton-neutron decomposition of the Hamiltonian, and present evidence for the viability of such a decomposition to reduce the size of the model space, through three different studies. The first is a series of calculations of proton-neutron entanglement entropy, a relatively novel approach in shell model calculations. Entanglement entropy measures the distribution of wavefunction coefficients, and thus the viability of truncation of a model space. These calculations study the strength and origin of the isospin dependence of the proton-neutron entanglement entropy. The second is a toy model that attempts to reproduce the entanglement entropy properties of realistic nuclear calculations. The third is a strength function decomposition of exact wavefunctions in an explicit proton-neutron formalism. Finally, I discuss a code to calculate nuclear wave-functions by a coupling of proton and neutron wave-functions which are calculated beforehand by an existing interacting shell model code. Results and convergence properties of this code are provided and discussed.