The first part of this thesis concerned with the solution structure determination problem. Whereas many biomacromolecules, such as proteins, can be adequately characterized by a single conformation in solution, numerous other important molecules (e.g., nucleic acids, carbohydrates, and polypeptides) exhibit conformational isomerism and disorder. For these molecules, the term "structure" does not correspond toa single conformation but rather to an ensemble of conformations. Given a molecular model and experimental data, the goal of the structure determination problem is to solve for an ensemble of conformations that is consistent with the data. Traditional computational procedures such as simulated annealing, however, are not guaranteed to generate a unique ensemble. The computed ensemble is often simply dependent on the user-specific protocol employed to generate it.

As an alternative, a numerical method for determining the conformational structure of macromolecules is developed and applied to idealized biomacromolecules in solution. The procedure generates unique, maximum entropy conformational ensembles that reproduce thermodynamic properties of the macromolecule (mean energy and heat capacity) in addition to the target experimental data. As an evaluation of its utility in structure determination, the method is applied to a homopolymer and a heteropolymer model of a three-helix bundle protein. It is demonstrated that the procedure performs successfully at various thermodynamic state points, including the ordered globule, disordered globule, and random coil states.

In the second part of this thesis, a molecular model is developed and used to investigatehe properties of anionic glycosaminoglycan (GAG) molecules. GAGs are critically important to the structure and biomechanical properties of articular cartilage, an avascular tissue that provides alow-friction, protective lining to the ends of contacting bones during join locomotion. The tissue consists predominantly of two types of macromolecules, collagen and aggrecan. Aggrecan cosists of a linear protein backbone with a high mass fraction of covalently attached chondroitin sulfate (CS) GAGs, which endow cartilage with its high compressive modulus via osmotic action. During the onset and progression ofosteoarthritis, a debilitating joint disease that affects millions in the US alone, the chemical composition of CS (sulfate type, sulfate pattern, and molecular weight) changes, concomitantly with alterations inthe biomechanical properties of cartilage.

For this reason, it is of primary biological interest to understand the effects of CS chemical composition on itsconformation, titration behavior, and osmotic pressure. To enable the investigation ofthese properties, a coarse-grained model of CS is developed. Systematically derived from an all-atom description, the model enables the atomisticbased simulation of large-scale macromolecular assemblies relevant to cartilage biomechanics. Extensive comparison with experimental data demonstrates thathis computationally efficient model is also quantitatively predictive, despite the absence of any adjustable parameters. 4-sulfation of CS is found to significantly increase the intrinsic stiffness of CS, as measured by the characteristic ratio and persistence length in the limit of high ionic strength. Average sulfate density is found to increase CS stiffness at finite ionic strength due to electrostatic interactions that tend to stiffen the chain backbone. Sulfation type and pattern (the statistical distribution of sulfates along a CS chain) are not found to influence the osmotic pressure, which is found to be sensitive primarily to the mean volumetric fixed charge density.