Monoclonal antibodies (mAbs) are key therapeutic tools used to treat many diseases, but their full potential is limited by formulation challenges. A major obstacle is creating stable, high concentration antibody solutions with low enough viscosity for subcutaneous injection. Although mAbs share a common Y shaped structure, small sequence differences can strongly affect their solution behavior, influencing both structural properties and collective dynamics. A central challenge is predicting how an antibody’s amino acid sequence and resulting surface charge distribution determine macroscopic properties such as viscosity. Solving this requires methods that connect molecular scale structure to measurable solution behavior. In a recent study (Camerin et al., "Electrostatics and viscosity are strongly linked in concentrated antibody solutions", PNAS 2025), we advanced this goal by developing a bead based computational model that incorporates heterogeneous charge distributions. This coarse grained approach reproduced key experimental observables—including structure factors and viscosity—across wide ranges of concentration and ionic strength, showing that physically grounded models can reliably predict antibody solution behavior. This proposal builds on that framework to study QAX576, a monoclonal antibody provided by Novartis. QAX576 (IgG1/k) targets human interleukin 13 and consists of two heavy chains (450 amino acids) and two light chains (216 amino acids). It is currently being characterized experimentally through large scale facility measurements and laboratory techniques within the LINXS program “Antibodies in Solutions,” which brings together academic groups, NIST, and pharmaceutical partners. Early measurements show unusual solution behavior, including thermodynamic boundaries whose molecular origins remain unclear. To investigate this, we will use a multiscale computational strategy that captures both the internal structure of QAX and its collective behavior in solution. First, atomistic molecular dynamics simulations will characterize the antibody’s conformational properties and provide structural input for coarse grained modeling. Next, we will build an amino acid–level coarse grained model. Constant pH Monte Carlo simulations with titration moves will determine the effective charge distribution under relevant conditions. The resulting amino-acid model will enable ensemble simulations aimed at systematically investigating how structural features and charge heterogeneity influence the structural organization of antibody solutions. Finally, we will employ a higher level colloidal bead model to compute dynamical observables, including zero shear viscosity—a key parameter for designing injectable high concentration formulations. Successful completion of this project will deliver a multiscale framework that links molecular level charge heterogeneity to macroscopic behavior. It will also provide insights valuable for antibody design and for developing therapeutic formulations with targeted physical properties.