There is a critical need to understand how tissue culture stimulation affects tissue construct development and function, with the ultimate goal of eliminating resource-intensive trial-and-error screening. Our goal is to develop predictive assessments of the in vivo performance of biomaterials so that a more rational approach based on a bottom-up modeling toolkit is used to guide the preparation of the required biomaterials. This new predictive approach would save time, animals, costs and accelerate the translation of such repair and regenerative systems. An important feature of our proposed approach is the direct integration of modeling and experimentation at multiple length scales, and the use of hierarchical material architectures across length scales, to reach enhanced material function. Our hypothesis is that predictions of biomaterials performance can be attained by the combined use of suitable experimental models to cover polymer features (chemistry, molecular weight), processing (fiber mechanical properties, hierarchical structure, degradation rate) and modeling at different length scales of materials structural hierarchy (from chemical to macroscopic). We have selected load bearing applications as the focus due to the generic needs in this field, such as for the anterior cruciate ligament, rotator cuff, bladder slings, hernia meshes, blood vessels, nerve guides and other tissues. Two well studied degradable polymer systems, silks and collagen, will be used for the experimental studies and model building, as they are directly amendable to highly controlled preparations and processing and cover a range of mechanical properties and degradation rates. In all cases, we build upon our extensive prior studies with these protein-based biomaterials, as well as developing hierarchical models of protein structure and function.