Department of Integrative Structural and Computational Biology, The Scripps Research Institute, La Jolla, CA 92037, USA.
Department of Bioengineering and Therapeutic Science and California Institute for Quantitative Biology, University of California, San Francisco, San Francisco, CA, USA.
Structure. 2019 Feb 5;27(2):344-358.e3. doi: 10.1016/j.str.2018.10.003. Epub 2018 Nov 15.
There does not currently exist a standardized indicator of how well cryo-EM-derived models represent the density from which they were generated. We present a straightforward methodology that utilizes freely available tools to generate a suite of independent models and to evaluate their convergence in an EM density. These analyses provide both a quantitative and qualitative assessment of the precision of the models and their representation of the density, respectively, while concurrently providing a platform for assessing both global and local EM map quality. We further use standardized datasets to provide an expected deviation within a suite of models refined against EM maps reported to be at 5 Å resolution or better. Associating multiple atomic models with a deposited EM map provides a rapid and accessible reporter of convergence, a strong indicator of highly resolved molecular detail, and is an important step toward an FSC-independent assessment of map and model quality.
目前还没有一个标准化的指标来衡量冷冻电镜(cryo-EM)衍生模型对其原始密度的代表性。我们提出了一种简单的方法,利用免费的工具生成一组独立的模型,并评估它们在电镜密度中的收敛性。这些分析分别为模型的精度和对密度的代表性提供了定量和定性的评估,同时为评估全局和局部电镜图质量提供了一个平台。我们进一步使用标准化数据集,为一组针对报告分辨率为 5Å 或更高的电镜图进行细化的模型提供预期偏差。将多个原子模型与已发表的电镜图相关联,为模型的收敛性提供了一个快速而易于访问的报告,这是对高分辨率分子细节的强烈指示,也是朝着独立于 FSC 评估图和模型质量的重要步骤。