High-throughput assays that produce hundreds of measurements per sample are powerful tools for quantifying cell–material interactions. With advances in automation and miniaturization in material fabrication, hundreds of biomaterial samples can be rapidly produced, which can then be characterized using these assays. However, the resulting deluge of data can be overwhelming. To the rescue are computational methods that are well suited to these problems. Machine learning techniques provide a vast array of tools to make predictions about cell–material interactions and to find patterns in cellular responses. Computational simulations allow researchers to pose and test hypotheses and perform experiments in silico. This review describes approaches from these two domains that can be brought to bear on the problem of analyzing biomaterial screening data.

每个样品可进行数百次测量的高通量分析是定量细胞与材料相互作用的有力工具。随着自动化和材料制造小型化的进步，可以快速生产数百种生物材料样品，然后可以使用这些测定法对其进行表征。但是，由此产生的大量数据可能是压倒性的。抢救这些方法非常适合这些问题。机器学习技术提供了大量工具，可以预测细胞与材料之间的相互作用，并找到细胞反应的模式。计算机模拟允许研究人员提出并证实假说，并进行实验，在硅片。这篇综述描述了来自这两个领域的方法，这些方法可以用于分析生物材料筛选数据的问题。