Microbiome engineering aims to manipulate, control, and design community-level properties through targeted interventions of existing microbial communities or the construction of new synthetic consortia. These efforts often lead to unexpected or undesirable outcomes because of highly complex input-output relationships that are primarily ascribable to adaptive responses of interspecies interactions to perturbation. Therefore, accurate prediction of microbial interaction networks and context-specific organization will aid success in future microbiome engineering efforts. Here, we review state-of-the-art modeling approaches to evaluate their scope of prediction as in silico tools for microbiome design. We highlight the utility of advanced models for predicting context-dependent interactions, multi-omics data integration, and combined use of complementary modeling and computational tools for enhanced prediction and eventual facilitation of in silico microbiome design.

微生物组工程旨在通过对现有微生物群落的有针对性的干预或新的合成财团的建设来操纵，控制和设计群落级的特性。由于高度复杂的投入产出关系（主要归因于种间相互作用对摄动的适应性反应），这些努力通常会导致意想不到的或不良的结果。因此，准确预测微生物相互作用网络和特定环境的组织将有助于未来微生物组工程的成功。在这里，我们回顾了最先进的建模方法，以评估其预测范围，如计算机模拟微生物组设计工具。我们重点介绍了高级模型在预测上下文相关的相互作用，多组学数据集成以及互补模型和计算工具的组合使用方面的实用性，以增强计算机微生物组设计的预测和最终的便利性。