Mechanistic models of how single cells respond to different perturbagens can help integrate disparate big data sets or predict response to varied drug combinations. However, the construction and simulation of such models have proved challenging. Our lab previously constructed one of the largest mechanistic models for single mammalian cell regulation of proliferation and death (774 species, 141 genes, 8 ligands, 2400 reactions). However, this, as many other large-scale models, was written using licensed software (MATLAB) with intricate programming structure, impeding alteration, expansion, and sharing. Here, we generated a new foundation for this model, which includes a python-based creation and simulation pipeline converting a few structured text files into an SBML-compatible format. This new open-source model (named SPARCED) is high-performance- and cloud-computing compatible and enables the study of virtual cell population responses at the single-cell level. We applied this new model to a subset of the LINCS MCF10A Data Cube, which observed that IFNγ acts as an anti-proliferative factor, but the reasons why were unknown. After expanding the SPARCED model with an IFNγ signaling module (to 950 species, 150 genes, 9 ligands, 2500 reactions), we ran stochastic single-cell simulations for two different putative crosstalk mechanisms and looked at the number of cycling cells in each case. Our model-based analysis suggested, and experiments support that these observations are better explained by IFNγ-induced SOCS1 expression sequestering activated EGF receptors, thereby downregulating AKT activity, as opposed to direct IFNγ-induced upregulation of p21 expression. This work forms a foundation for increased mechanistic model-based data integration on a single-cell level, an important building block for clinically predictive mechanistic models.

中文翻译：

---

用于单细胞增殖和死亡信号的大规模机制模型的可扩展、开源实现

单细胞如何响应不同扰动因子的机制模型可以帮助整合不同的大数据集或预测对不同药物组合的响应。然而，这种模型的构建和模拟已被证明具有挑战性。我们的实验室之前构建了最大的单个哺乳动物细胞增殖和死亡调控机制模型之一（774 个物种、141 个基因、8 个配体、2400 个反应）。然而，这与许多其他大型模型一样，是使用许可软件（MATLAB）编写的，具有复杂的编程结构，阻碍了更改、扩展和共享。在这里，我们为这个模型生成了一个新的基础，其中包括一个基于 python 的创建和模拟管道，将一些结构化的文本文件转换为与 SBML 兼容的格式。这种新的开源模型（名为 SPARCED）与高性能和云计算兼容，可以在单细胞水平上研究虚拟细胞群反应。我们将这个新模型应用于 LINCS MCF10A 数据立方体的一个子集，它观察到 IFNγ 作为一种抗增殖因子，但原因未知。在使用 IFNγ 信号模块扩展 SPARCED 模型后（扩展到 950 个物种、150 个基因、9 个配体、2500 个反应），我们对两种不同的假定串扰机制进行了随机单细胞模拟，并观察了每种情况下循环细胞的数量。我们基于模型的分析表明，并且实验支持这些观察结果可以通过 IFNγ 诱导的 SOCS1 表达隔离激活的 EGF 受体更好地解释，从而下调 AKT 活性，与直接 IFNγ 诱导的 p21 表达上调相反。这项工作为在单细胞水平上增加基于机械模型的数据集成奠定了基础，这是临床预测机械模型的重要组成部分。