Dataset Information

Dilated-DenseNet For Macromolecule Classification In Cryo-electron Tomography.

ABSTRACT: Cryo-electron tomography (cryo-ET) combined with subtomogram averaging (STA) is a unique technique in revealing macromolecule structures in their near-native state. However, due to the macromolecular structural heterogeneity, low signal-to-noise-ratio (SNR) and anisotropic resolution in the tomogram, macromolecule classification, a critical step of STA, remains a great challenge. In this paper, we propose a novel convolution neural network, named 3D-Dilated-DenseNet, to improve the performance of macromolecule classification in STA. The proposed 3D-Dilated-DenseNet is challenged by the synthetic dataset in the SHREC contest and the experimental dataset, and compared with the SHREC-CNN (the state-of-the-art CNN model in the SHREC contest) and the baseline 3D-DenseNet. The results showed that 3D-Dilated-DenseNet significantly outperformed 3D-DenseNet but 3D-DenseNet is well above SHREC-CNN. Moreover, in order to further demonstrate the validity of dilated convolution in the classification task, we visualized the feature map of 3D-Dilated-DenseNet and 3D-DenseNet. Dilated convolution extracts a much more representative feature map.

SUBMITTER: Gao S

PROVIDER: S-EPMC8046028 | biostudies-literature | 2020 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Dilated-DenseNet For Macromolecule Classification In Cryo-electron Tomography.

Gao Shan S Han Renmin R Zeng Xiangrui X Cui Xuefeng X Liu Zhiyong Z Xu Min M Zhang Fa F

Bioinformatics research and applications : ... international symposium, ISBRA ... proceedings. ISBRA (Conference) 20200818

Cryo-electron tomography (cryo-ET) combined with subtomogram averaging (STA) is a unique technique in revealing macromolecule structures in their near-native state. However, due to the macromolecular structural heterogeneity, low signal-to-noise-ratio (SNR) and anisotropic resolution in the tomogram, macromolecule classification, a critical step of STA, remains a great challenge. In this paper, we propose a novel convolution neural network, named 3D-Dilated-DenseNet, to improve the performance o ...[more]

PMID: 33860285

Similar Datasets

Project description:Cryo-electron tomography (Cryo-ET) is an emerging technology for three-dimensional (3D) visualization of macromolecular structures in the near-native state. To recover structures of macromolecules, millions of diverse macromolecules captured in tomograms should be accurately classified into structurally homogeneous subsets. Although existing supervised deep learning-based methods have improved classification accuracy, such trained models have limited ability to classify novel macromolecules that are unseen in the training stage. To adapt the trained model to the macromolecule classification of a novel class, massive labeled macromolecules of the novel class are needed. However, data labeling is very time-consuming and labor-intensive. In this work, we propose a novel few-shot learning method for the classification of novel macromolecules (named FSCC). A two-stage training strategy is designed in FSCC to enhance the generalization ability of the model to novel macromolecules. First, FSCC uses contrastive learning to pre-train the model on a sufficient number of labeled macromolecules. Second, FSCC uses distribution calibration to re-train the classifier, enabling the model to classify macromolecules of novel classes (unseen class in the pre-training). Distribution calibration transfers learned knowledge in the pre-training stage to novel macromolecules with limited labeled macromolecules of novel class. Experiments were performed on both synthetic and real datasets. On the synthetic datasets, compared with the state-of-the-art (SOTA) method based on supervised deep learning, FSCC achieves competitive performance. To achieve such performance, FSCC only needs five labeled macromolecules per novel class. However, the SOTA method needs 1100 ∼ 1500 labeled macromolecules per novel class. On the real datasets, FSCC improves the accuracy by 5% ∼ 16% when compared to the baseline model. These demonstrate good generalization ability of contrastive learning and calibration distribution to classify novel macromolecules with very few labeled macromolecules.

Project description:BackgroundDespite recent advances in cellular cryo-electron tomography (CET), developing automated tools for macromolecule identification in submolecular resolution remains challenging due to the lack of annotated data and high structural complexities. To date, the extent of the deep learning methods constructed for this problem is limited to conventional Convolutional Neural Networks (CNNs). Identifying macromolecules of different types and sizes is a tedious and time-consuming task. In this paper, we employ a capsule-based architecture to automate the task of macromolecule identification, that we refer to as 3D-UCaps. In particular, the architecture is composed of three components: feature extractor, capsule encoder, and CNN decoder. The feature extractor converts voxel intensities of input sub-tomograms to activities of local features. The encoder is a 3D Capsule Network (CapsNet) that takes local features to generate a low-dimensional representation of the input. Then, a 3D CNN decoder reconstructs the sub-tomograms from the given representation by upsampling.ResultsWe performed binary and multi-class localization and identification tasks on synthetic and experimental data. We observed that the 3D-UNet and the 3D-UCaps had an [Formula: see text]score mostly above 60% and 70%, respectively, on the test data. In both network architectures, we observed degradation of at least 40% in the [Formula: see text]-score when identifying very small particles (PDB entry 3GL1) compared to a large particle (PDB entry 4D8Q). In the multi-class identification task of experimental data, 3D-UCaps had an [Formula: see text]-score of 91% on the test data in contrast to 64% of the 3D-UNet. The better [Formula: see text]-score of 3D-UCaps compared to 3D-UNet is obtained by a higher precision score. We speculate this to be due to the capsule network employed in the encoder. To study the effect of the CapsNet-based encoder architecture further, we performed an ablation study and perceived that the [Formula: see text]-score is boosted as network depth is increased which is in contrast to the previously reported results for the 3D-UNet. To present a reproducible work, source code, trained models, data as well as visualization results are made publicly available.ConclusionQuantitative and qualitative results show that 3D-UCaps successfully perform various downstream tasks including identification and localization of macromolecules and can at least compete with CNN architectures for this task. Given that the capsule layers extract both the existence probability and the orientation of the molecules, this architecture has the potential to lead to representations of the data that are better interpretable than those of 3D-UNet.

Dataset Information

Dilated-DenseNet For Macromolecule Classification In Cryo-electron Tomography.

Publications

Dilated-DenseNet For Macromolecule Classification In Cryo-electron Tomography.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets