BeReTa: a systematic method for identifying target transcriptional regulators to enhance microbial production of chemicals
BeReTa: a systematic method for identifying target transcriptional regulators to enhance microbial production of chemicals
复制标题
DOI：
10.1093/bioinformatics/btw557
复制
发表时间：
2017-01-01
期刊：
影响因子：
5.8
通讯作者：
Kim, Byung-Gee
中科院分区：
文献类型：
--
作者：
Kim, Minsuk;Sun, Gwanggyu;Kim, Byung-Gee
动机：调控代谢过程的调控电路是发展微生物细胞工厂的关键步骤。尽管在硅应变设计算法的流行，他们中的大多数都不能预测所需的修改调节网络。虽然一些算法可以预测转录调控（TR）操作的相关靶标，但由于它们高度依赖于综合代谢/调控模型的可用性，它们的可靠性和适用性有限。结果：我们提出了BeReTa （Beneficial Regulator Targeting）算法，这是一种利用非集成网络模型对TR操纵目标进行优先排序的新算法。BeReTa通过评估相互作用的调节强度和反应的有益效果来确定TR操纵目标，并随后为TR分配有益分数。我们证明BeReTa可以预测已知的和新的TR操纵靶标，以增强大肠杆菌中各种化学物质的生产。此外，通过对彩色链霉菌生产抗生素的案例研究，我们成功地证明了它在研究较少的生物体中的广泛适用性。据我们所知，BeReTa是第一个专门用于预测TR操作目标的应变设计算法。
Motivation: Modulation of regulatory circuits governing the metabolic processes is a crucial step for developing microbial cell factories. Despite the prevalence of in silico strain design algorithms, most of them are not capable of predicting required modifications in regulatory networks. Although a few algorithms may predict relevant targets for transcriptional regulator (TR) manipulations, they have limited reliability and applicability due to their high dependency on the availability of integrated metabolic/regulatory models.Results: We present BeReTa (Beneficial Regulator Targeting), a new algorithm for prioritization of TR manipulation targets, which makes use of unintegrated network models. BeReTa identifies TR manipulation targets by evaluating regulatory strengths of interactions and beneficial effects of reactions, and subsequently assigning beneficial scores for the TRs. We demonstrate that BeReTa can predict both known and novel TR manipulation targets for enhanced production of various chemicals in Escherichia coli. Furthermore, through a case study of antibiotics production in Streptomyces coelicolor, we successfully demonstrate its wide applicability to even less-studied organisms. To the best of our knowledge, BeReTa is the first strain design algorithm exclusively designed for predicting TR manipulation targets.