Marriage of High-Throughput Gradient Surface Generation with Statistical Learning for the Rational Design of Functionalized Biomaterials.

Abstract

Functional biomaterial has already become an important aspect in modern therapeutics, yet the design of novel multi-functional biomaterial is still a challenging task nowadays. When several biofunctional components are present, the complexity that arises from their combinations and interactions would lead to tedious trial-and-error screening. In this work, we present a novel strategy of biomaterial rational design through the marriage of gradient surface generation with statistical learning. Not only parameter combinations can be screened in a high-throughput fashion, the optimal conditions beyond the experimentally tested range can also be extrapolated from the models. We have demonstrated the power of our strategy in rationally designing an unprecedented ternary functionalized surface for orthopedic implant, with optimal osteogenic, angiogenic and neurogenic activities, and have confirmed its optimality in vitro and the best osteointegration promotion in vivo. The presented strategy is expected to open up new possibilities in the rational design of biomaterials. This article is protected by copyright. All rights reserved.