Dudola Dániel, Kovács Bertalan, Gáspári Zoltán
Faculty of Information Technology and Bionics, Pázmány Péter Catholic University , 1083 Budapest, Práter u. 50/A, Hungary.
J Chem Inf Model. 2017 Aug 28;57(8):1728-1734. doi: 10.1021/acs.jcim.7b00066. Epub 2017 Aug 3.
Ensemble-based models of protein structure and dynamics reflecting experimental parameters are increasingly used to obtain deeper understanding of the role of dynamics in protein function. Such ensembles differ substantially from those routinely deposited in the PDB and, consequently, require specialized validation and analysis methodology. Here we describe our completely rewritten online validation tool, CoNSEnsX, that offers a standardized way to assess the correspondence of such ensembles to experimental NMR parameters. The server provides a novel selection feature allowing a user-selectable set and weights of different parameters to be considered. This also offers an approximation of potential overfitting, namely, whether the number of conformers necessary to reflect experimental parameters can be reduced in the ensemble provided. The CoNSEnsX webserver is available at consensx.itk.ppke.hu . The corresponding Python source code is freely available on GitHub ( github.com/PPKE-Bioinf/consensx.itk.ppke.hu ).
反映实验参数的基于集成的蛋白质结构和动力学模型越来越多地被用于更深入地理解动力学在蛋白质功能中的作用。此类集成与通常存于蛋白质数据银行（PDB）中的集成有很大不同，因此需要专门的验证和分析方法。在此，我们描述了我们完全重写的在线验证工具CoNSEnsX，它提供了一种标准化方法来评估此类集成与实验核磁共振（NMR）参数的对应性。该服务器提供了一种新颖的选择功能，允许考虑用户可选择的不同参数集和权重。这也提供了潜在过拟合的近似值，即在所提供的集成中反映实验参数所需的构象数是否可以减少。CoNSEnsX网络服务器可在consensx.itk.ppke.hu上获取。相应的Python源代码可在GitHub（github.com/PPKE - Bioinf/consensx.itk.ppke.hu）上免费获取。
