Computational Biology Lab, MBG-CSIC, Spanish National Research Council, 36143 Pontevedra, Spain.
Department of Mathematics and CITIC, Universidade da Coruña, 15071 A Coruña, Spain.
ACS Synth Biol. 2023 Oct 20;12(10):2865-2876. doi: 10.1021/acssynbio.3c00033. Epub 2023 Oct 9.
Microorganisms (mainly bacteria and yeast) are frequently used as hosts for genetic constructs in synthetic biology applications. Molecular noise might have a significant effect on the dynamics of gene regulation in microbial cells, mainly attributed to the low copy numbers of mRNA species involved. However, the inclusion of molecular noise in the automated design of biocircuits is not a common practice due to the computational burden linked to the chemical master equation describing the dynamics of stochastic gene regulatory circuits. Here, we address the automated design of synthetic gene circuits under the effect of molecular noise combining a mixed integer nonlinear global optimization method with a partial integro-differential equation model describing the evolution of stochastic gene regulatory systems that approximates very efficiently the chemical master equation. We demonstrate the performance of the proposed methodology through a number of examples of relevance in synthetic biology, including different bimodal stochastic gene switches, robust stochastic oscillators, and circuits capable of achieving biochemical adaptation under noise.
微生物（主要是细菌和酵母）经常被用作合成生物学应用中基因构建体的宿主。分子噪声可能对微生物细胞中基因调控的动力学产生重大影响，主要归因于涉及的 mRNA 种类的低拷贝数。然而，由于描述随机基因调控电路动力学的化学主方程所带来的计算负担，将分子噪声纳入生物电路的自动设计中并不是一种常见做法。在这里，我们结合混合整数非线性全局优化方法和描述随机基因调控系统演化的偏积分微分方程模型来解决分子噪声影响下的合成基因电路的自动设计问题，该模型能够非常有效地逼近化学主方程。我们通过一些在合成生物学中具有重要意义的例子展示了所提出方法的性能，包括不同的双模态随机基因开关、鲁棒随机振荡器，以及能够在噪声下实现生化适应的电路。
