Prediction of least energy conformation of a protein from its primary structure (chain of amino acids) is an optimization problem associated with a large complex energy landscape. In this study, a simple 2D hydrophobic–hydrophilic model was used to model the protein sequence, which allows the fast and efficient design of genetic algorithm-based protein structure prediction approach. The neighborhood search strategy is integrated into the genetic operator. The neighborhood search guides the genetic operator to regions in the computational space with good solutions. To prevent convergence to local optima, the proposed method employs crowding-based parent replacement strategy, which improves the performance of the algorithm and the ability to deal with multiple numbers of solutions. The proposed algorithm was tested with a standard benchmark of HP sequences and comparative results demonstrate that the proposed system beats most of the evolutionary algorithms for seven sequences. It finds the best energy for a sequence of length [Formula: see text], [Formula: see text], [Formula: see text], [Formula: see text], [Formula: see text], [Formula: see text] and [Formula: see text].

中文翻译：

---

HPS_PSP：蛋白质结构预测的高性能系统

从蛋白质的一级结构（氨基酸链）预测蛋白质的最小能量构象是与大型复杂能量景观相关的优化问题。在这项研究中，使用简单的 2D 疏水-亲水模型对蛋白质序列进行建模，从而可以快速有效地设计基于遗传算法的蛋白质结构预测方法。邻域搜索策略被集成到遗传算子中。邻域搜索将遗传算子引导到计算空间中具有良好解的区域。为了防止收敛到局部最优，该方法采用了基于拥挤的父代替换策略，提高了算法的性能和处理多个解的能力。所提出的算法使用 HP 序列的标准基准进行了测试，比较结果表明，所提出的系统击败了七个序列的大多数进化算法。它为长度序列找到最佳能量 [公式：见文本]、[公式：见文本]、[公式：见文本]、[公式：见文本]、[公式：见文本]、[公式：见文本] 和 [公式：见正文]。