Posgrado en Ingeniería Eléctrica, Universidad Nacional Autónoma de México, Cd.Universitaria, C.P.04510, Mexico City, Mexico.
Department of Computer Science, Instituto de Investigaciones en Matemáticas Aplicadas y en Sistemas, Universidad Nacional Autónoma de México, Mexico City 04510, Mexico.
J Struct Biol. 2021 Sep;213(3):107771. doi: 10.1016/j.jsb.2021.107771. Epub 2021 Jul 26.
The quality of a 3D map produced by the single-particle analysis method is highly dependent on an accurate assignment of orientations to the many experimental images. However, the problem's complexity implies the presence of several local minima in the optimized goal functions. Consequently, validation methods to confirm the angular assignment are very useful to yield higher-resolution 3D maps. In this work, we present a graph-signal-processing-based methodology that analyzes the correlation landscape as a function of the orientation, an approach allowing the estimation of the assigned orientations' reliability. Using this method, we may identify low-reliability images that probably incorrectly contribute to the final 3D reconstruction.
由单颗粒分析法产生的 3D 图谱的质量高度依赖于对大量实验图像的取向的精确分配。然而，问题的复杂性意味着在优化的目标函数中存在几个局部最小值。因此，验证方法来确认角度分配是非常有用的，可以得到更高分辨率的 3D 图谱。在这项工作中，我们提出了一种基于图信号处理的方法，该方法分析了作为取向函数的相关景观，这是一种可以估计所分配取向的可靠性的方法。使用这种方法，我们可以识别可能错误地对最终 3D 重建做出贡献的低可靠性图像。
