The translation of proteins in cellular environments and the diffusion coefficientsof particles play crucial roles in biological processes. However, experimental investigations to determine these coefficients are challenging, requiring time-consuming procedures, expensive high purity samples, and specialized instrumentation. To overcome these issues, computational estimations of diffusion coefficients offer a promising solution. To achieve this computationally, all-atom molecular dynamics simulations are used. These calculations are computationally demanding, requiring long-time simulations of biomolecules with thousands of atoms in aqueous environments. To address both experimental and computational challenges, a new approach is taken, focusing on the fundamental building blocks of protein and polysaccharide structures and how they collectively produce the properties and behavior of larger assemblies. By linking the local chemistry of a solvent-exposed patch of a molecule to its contribution to the overall hydrodynamic radius, an accurate determination of the experimentally comparable diffusion coefficient is achieved. The resulting predictions for diffusion coefficients in various protein structures are comparable to state-of-the-art statistical methods, which may heavily rely on limited training data. This research not only applies to large protein structures but also provides predictive insights into the dynamics of oligosaccharides and larger dendrimers. By understanding the fundamental aspects of molecular diffusion through this approach, the study opens up possibilities for bypassing experimental complexities and enhancing our understanding of biological processes at a deeper level.