Multivariate statistical analysis of three-dimensional cross-bridge motifs in insect flight muscle
Multivariate statistical analysis of three-dimensional cross-bridge motifs in insect flight muscle
复制标题
DOI：
10.1016/s0304-3991(99)00035-2
复制
发表时间：
1999-07-01
期刊：
影响因子：
2.2
通讯作者：
Taylor, KA
中科院分区：
文献类型：
--
作者：
Winkler, H;Taylor, KA
多元统计分析已被应用于识别昆虫飞行肌肉断层图中相似的三维图案的问题。该方法是将该技术广泛应用于二维图像的扩展。由于灯丝晶格中非常规则的基序排列，3d情况施加的额外自由度已经减少到围绕灯丝轴的平移对齐和180度旋转。通过使用对对齐参数不变的多元统计分析的派生函数，解决了寻找对齐无偏参考的问题。层次上升分类是一种无监督分类方法。结果表明，类平均的信噪比有所提高，密度分布与已知的丝格内肌凝蛋白头和肌动蛋白单体的数量保持一致。(C) 1999 Elsevier Science B.V.版权所有
Multivariate statistical analysis has been applied to the problem of identifying similar 3-D motifs within tomograms of insect flight muscle. The method is an extension of the widely used application of this technique to 2-D images. The additional degrees of freedom that the 3-D case imposes has been reduced to a translational alignment and a 180 degrees rotation about the filament axis because of the very regular motif arrangement in the filament lattice. The problem of finding unbiased references for the alignment has been addressed by using derived functions for the multivariate statistical analysis that are invariant to the alignment parameters. Hierarchical ascendant classification has been explored as an unsupervised classification method. The results show improved signal to noise ratio in the class averages with retention of density distributions consistent with the known numbers of myosin heads and actin monomers present within the filament lattice. (C) 1999 Elsevier Science B.V. All rights reserved.