Directed evolution methods mimic in vitro Darwinian evolution, inducing random mutations and selective pressure in genes to obtain proteins with enhanced characteristics. These techniques are developed using trial-and-error testing at an experimental level with a high degree of uncertainty. Therefore, in silico modeling of directed evolution is required to support experimental assays. Several in silico approaches have reproduced directed evolution, using statistical, thermodynamic, and kinetic models in an attempt to recreate experimental conditions. Likewise, optimization techniques using heuristic models have been used to understand and find the best scenarios of directed evolution. Our study uses an in silico model named HeurIstics DirecteD EvolutioN, which is based on a genetic algorithm designed to generate chimeric libraries from 2 parental genes, cry11Aa and cry11Ba, of Bacillus thuringiensis. These genes encode crystal-shaped δ-endotoxins with 3 conserved domains. Cry11 toxins are of biotechnological interest because they have shown to be effective as biopesticides for disease-spreading vectors. With our heuristic model, we considered experimental parameters such as DNA fragmentation length, number of generations or simulation cycles, and mutation rate, to get characteristics of Cry11 chimeric libraries such as percentage of population identity, truncation of variants obtained from the presence of internal stop codons, percentage of thermodynamic diversity, and stability of variants. Our study allowed us to focus on experimental conditions that may be useful for the design of in vitro and in silico experiments of directed evolution with Cry toxins of 3 conserved domains. Furthermore, we obtained in silico libraries of Cry11 variants, in which structural characteristics of wild Cry families were observed in a review of a sample of in silico sequences. We consider that future studies could use our in silico libraries and heuristic computational models, as the one suggested here, to support in vitro experiments of directed evolution.

定向进化方法模拟体外达尔文进化，诱导基因中的随机突变和选择性压力以获得具有增强特性的蛋白质。这些技术是通过反复试验在不确定的实验水平上开发的。因此，需要进行定向进化的计算机模拟以支持实验测定。几种计算机模拟方法已使用统计，热力学和动力学模型重现了定向进化，以尝试重建实验条件。同样，使用启发式模型的优化技术已被用来理解和找到定向进化的最佳方案。我们的研究使用了一个名为HeurIstics DirecteD EvolutioN的计算机模型，cry11Aa和cry11Ba的苏云金芽孢杆菌。这些基因编码具有3个保守结构域的晶体形δ-内毒素。Cry11毒素具有生物技术意义，因为它们已被证明可有效用作疾病传播载体的生物农药。利用我们的启发式模型，我们考虑了诸如DNA片段长度，世代或模拟周期数以及突变率等实验参数，以获得Cry11的特征。嵌合文库，例如群体同一性百分比，从内部终止密码子存在获得的变体截短，热力学多样性百分比和变体稳定性。我们的研究使我们能够专注于可能对体外和计算机模拟设计的3个保守域的Cry毒素进行定向进化的实验条件。此外，我们获得了Cry11变体的计算机文库，其中在计算机硅序列样本的综述中观察到了野生Cry家族的结构特征。我们认为，如本文所述，未来的研究可以使用我们的计算机模拟库和启发式计算模型来支持定向进化的体外实验。