...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Knowledge-Based Systems >An improved memetic approach for protein structure prediction incorporating maximal hydrophobic core estimation concept
【24h】

An improved memetic approach for protein structure prediction incorporating maximal hydrophobic core estimation concept

机译：一种改进的蛋白质结构预测膜，包括最大疏水核心估计概念

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Protein Structure Prediction (PSP) from the primary amino acid sequence, even using a simplified Hydrophobic-Polar (HP) lattice model, continues to be extremely challenging. Finding an optimal conformation, even for a small sequence, by any of the currently known evolutionary approaches is computationally extensive and time consuming. Although Memetic Algorithms (MAs) have shown success in finding the optimal solution for PSP, no significant work on the incorporation of domain or problem specific knowledge into the search process to significantly improve their performance is reported. In this paper, we present an approach to incorporate such knowledge into the initial population to enhance the effectiveness of MA for PSP. The domain knowledge we propose to use is based on the concept of maximal 'core' formation by exploiting the fundamental property of the H residues to be at the core of the minimum energy optimal protein structure. A generic technique is proposed for estimating the maximal Hydrophobic core (H-core) in a protein sequence for 2D Square, 3D Cubic and a more complex and realistic 3D FCC (Face Centered Cubic) lattice models. Subsequently, the knowledge of this estimated core is incorporated in an MA. The experiments conducted using HP benchmark sequences for 2D Square, 3D Cubic and 3D FCC lattice models show that the proposed MA with the new core-based population initialization technique has superior performance to the existing methods in terms of convergence speed as well as minimal energy. (C) 2018 Elsevier B.V. All rights reserved.
机译：蛋白质结构预测（PSP）来自伯氨基酸序列，甚至使用简化的疏水性（HP）格子模型，仍然是极其具有挑战性的。通过任何当前已知的进化方法找到最佳构象，即使是小序列也在计算广泛且耗时。虽然Memet算法（MAS）在找到PSP的最佳解决方案方面已经取得了成功，但没有对域或问题的结合到搜索过程中没有重大工作，以显着提高它们的性能。在本文中，我们提出了一种将这些知识纳入初始群体的方法，以提高MA为PSP的有效性。我们建议使用的域名知识是基于最大“核心”形成的概念，通过利用H残留物的基本属性处于最小能量最佳蛋白质结构的核心。提出了一种普遍的技术，用于估计2D平方，3D立方和更复杂和更加复杂的3D FCC（面为中心立方）格式模型的蛋白质序列中的最大疏水芯（H核）。随后，将该估计核心的知识纳入MA。使用HP基准序列进行2D平方，3D立方和3D FCC格晶格模型进行的实验表明，具有新的基于核心的人口初始化技术的提出的MA对现有方法具有卓越的性能，以及在收敛速度以及最小的能量方面对现有方法具有卓越的性能。 （c）2018 Elsevier B.v.保留所有权利。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号