山东科学 ›› 2015, Vol. 28 ›› Issue (3): 96-103.doi: 10.3976/j.issn.1002-4026.2015.03.018

• 其他研究论文 • 上一篇    下一篇

蛋白质微秒级折叠过程的快速模拟

林云,张少龙*,张庆刚,张怿慈   

  1. 山东师范大学物理与电子科学学院,山东 济南 250014
  • 收稿日期:2015-03-23 出版日期:2015-06-20 发布日期:2015-06-20
  • 通信作者: 张少龙 E-mail:slzhang@sdnu.edu.cn
  • 作者简介:林云(1989-),男,硕士研究生,研究方向为分子动力学模拟。
  • 基金资助:
    国家自然科学基金(11274206);山东省自然科学基金(ZR2011HM048)

Rapid simulation of protein microsecond level folding process

LIN Yun, ZHANG Shaolong,ZHANG Qinggang, ZHANG Yici   

  1. School of Physics and Electronics,Shandong Normal University,Jinan 250014
  • Received:2015-03-23 Online:2015-06-20 Published:2015-06-20

摘要: 使用ff12SB力场和广义玻恩(GBNeck2)隐性水模型在GTX670 GPU上对4个蛋白质CLN025 (2ZEI)、MHA6(2I9M)、TrpCage (1L2Y)、Villin (3TRW)的微秒级折叠过程进行了分子动力学模拟,2ZEI的折叠时间和均方根偏差为5.94 μs和0.897 ,2I9M的折叠时间和均方根偏差为0.191 μs和1.142 ,Villin的折叠时间和均方根偏差为4.23 μs和1.37 ,TrpCage的折叠时间和均方根偏差为2.48 μs和0.63 。结果表明,蛋白质折叠模拟已经能够通过GPU计算在桌面计算机上实现。

关键词: 分子动力学, GPU计算, ff12SB力场, GBNeck2, 蛋白质折叠

Abstract: We performed molecular dynamics simulation for four proteins, CLN025 (2ZEI), MHA6 (2I9M), TrpCage (1L2Y) and Villin(3TRW) on GEX670GPU with AMBER ff12SB force field and Generalized Born (GBNeck2) implicit solvent model. Results show that folding time and RMSD compared to its crystal structure are 5.94 μs and 0.897  for protein 2ZEI, 0.191 μs and 1.142 for protein 219M, 4.23 μs and 1.37  for protein Villin, and 2.49 μs and 0.63  protein TrpCage.It demonstrates that microsecond level protein folding simulation can be implemented on a desktop workstation with GPU computing.

Key words: protein folding, GBNeck2, ff12SB force field, GPU computing, molecular dynamics

中图分类号: 

  • O641.12