Integrative computational framework to decipher the functions of shell proteins in biomineralization

Mollusk shells contain biominerals with remarkable mechanical properties enabled by a small fraction of embedded organic matrix proteins. However, the specific molecular functions of most shell proteins have remained elusive. Traditional genomics and functional studies are extremely laborious to identify key components. To address this, we developed an in-silico pipeline integrating protein structure modeling, molecular dynamics simulations, and machine learning to elucidate the critical ion protein interactions governing shell formation. Using the pearl oyster Pinctada fucata as a test case, our framework successfully recapitulated known protein functions and predicted roles of uncharacterized proteins to guide future experiments. Moreover, the pipeline’s modular design enables versatile applications for rapidly elucidating structure-function relationships in diverse biomineralization systems, complementing conventional wet-lab methods. Overall, this computational approach leverages automatic simulations and analytics to unlock molecular insights into shell protein ion dynamics, accelerating the discovery of key crystallization regulators for bioinspired materials design.