Intrinsically Disordered Proteins (IDPs) have an intricate involvement in intracellular phenomena. Their deep imprint on biological processes frequently implicates them in a variety of debilitating pathologies. They also pose complex scientific challenges to researchers due to their inadherence to our traditional understanding of protein functionality. Complementary research efforts between experimentalists and computational scientists are necessary to unlock our understanding of their physics.

This thesis primarily focuses on the current computational models that seek to shed light on the unique properties of IDPs. Models that generate an ensemble of IDP structures and those that codify the information in the ensembles to convenient forms are both necessary for a comparison with experimental data. Obtaining an agreement with the ensemble averaged properties from experiments is the first step towards unlocking the possibilities from computational models. The most attractive amongst them is an atomistic description of IDP ensembles.

Computational methods frequently resort to an implicit solvent assumption both to predict the conformations of IDPs and to interpret the generated data. The relaxation of a molecular solvent in implicit solvent models allows for their implementation with fewer computational resources. In the first part of this thesis, implicit solvent models of hydration and scattering are critically evaluated by isolating their central approximations. Implicit solvent models of hydration are shown to lack the necessary ability to discriminate between disparate conformations. Implicit solvent models of scattering analysis are shown to distort the information present in the system and blur our ability to detect the true merit of an ensemble generating method.

Explicit solvent models foundationally incorporate a molecular nature of the solvent and hence are logical counterparts to implicit solvent models. In the final part of this thesis, explicit water simulations are used to generate ensembles of polyampholytes that are then input into an explicit water scattering model. Models with differing philosophies are tested on their predictions for the same system in order to generate detailed insights that have the potential to benefit future studies. The recent updates in the atomistic modeling of both proteins and waters are shown to translate to a better agreement when comparing with experimental data.

The use of idealized mimics of IDPs also known as polyampholytes have allowed for the detailed studies in this thesis to be carried out. The ease of handling polyampholytes in computational studies promises to provide more foundational insights in the near future.