Inria Rennes: Inria Centre de Recherche Rennes Bretagne Atlantique, France.
Inria Rennes: Inria Centre de Recherche Rennes Bretagne Atlantique, France.
Comput Methods Programs Biomed. 2022 Oct;225:107017. doi: 10.1016/j.cmpb.2022.107017. Epub 2022 Jul 16.
Cryo electron tomography visualizes native cells at nanometer resolution, but analysis is challenged by noise and artifacts. Recently, supervised deep learning methods have been applied to decipher the 3D spatial distribution of macromolecules. However, in order to discover unknown objects, unsupervised classification techniques are necessary. In this paper, we provide an overview of unsupervised deep learning techniques, discuss the challenges to analyze cryo-ET data, and provide a proof-of-concept on real data.
We propose a weakly supervised subtomogram classification method based on transfer learning. We use a deep neural network to learn a clustering friendly representation able to capture 3D shapes in the presence of noise and artifacts. This representation is learned here from a synthetic data set.
We show that when applying k-means clustering given a learning-based representation, it becomes possible to satisfyingly classify real subtomograms according to structural similarity. It is worth noting that no manual annotation is used for performing classification.
We describe the advantages and limitations of our proof-of-concept and raise several perspectives for improving classification performance.
冷冻电子断层扫描以纳米分辨率可视化天然细胞，但分析受到噪声和伪影的挑战。最近，监督深度学习方法已被应用于破译大分子的 3D 空间分布。然而，为了发现未知对象，需要使用无监督分类技术。本文综述了无监督深度学习技术，讨论了分析冷冻电子断层扫描数据的挑战，并提供了真实数据的概念验证。
我们提出了一种基于迁移学习的弱监督子断层分类方法。我们使用深度神经网络学习一种聚类友好的表示形式，能够在存在噪声和伪影的情况下捕获 3D 形状。在此，我们从合成数据集学习该表示形式。
我们表明，在应用基于学习的表示形式的 k-均值聚类时，可以根据结构相似性令人满意地对真实的子断层进行分类。值得注意的是，分类过程没有使用任何手动注释。
我们描述了我们概念验证的优点和局限性，并提出了几种提高分类性能的观点。
