Active Learning to Classify Macromolecular Structures in situ for Less Supervision in Cryo-Electron Tomography

Abstract

Motivation: Cryo-Electron Tomography (cryo-ET) is a 3D bioimaging tool that visualizes the structural and spatial organization of macromolecules at a near-native state in single cells, which has broad applications in life science. However, the systematic structural recognition and recovery of macromolecules captured by cryo-ET are difficult due to high structural complexity and imaging limits. Deep learning based subtomogram classification have played critical roles for such tasks. As supervised approaches, however, their performance relies on sufficient and laborious annotation on a large training dataset. Results: To alleviate this major labeling burden, we proposed a Hybrid Active Learning (HAL) framework for querying subtomograms for labelling from a large unlabeled subtomogram pool. Firstly, HAL adopts uncertainty sampling to select the subtomograms that have the most uncertain predictions. Moreover, to mitigate the sampling bias caused by such strategy, a discriminator is introduced to judge if a certain subtomogram is labeled or unlabeled and subsequently the model queries the subtomogram that have higher probabilities to be unlabeled. Additionally, HAL introduces a subset sampling strategy to improve the diversity of the query set, so that the information overlap is decreased between the queried batches and the algorithmic efficiency is improved. Our experiments on subtomogram classification tasks using both simulated and real data demonstrate that we can achieve comparable testing performance (on average only 3% accuracy drop) by using less than 30% of the labeled subtomograms, which shows a very promising result for subtomogram classification task with limited labeling resources.

arXiv commentStatement on authorship changes: Dr. Eric Xing was an academic advisor of Mr. Haohan Wang. Dr. Xing was not directly involved in this work and has no direct interaction or collaboration with any other authors on this work. Therefore, Dr. Xing is removed from the author list according to his request. Mr. Zhenxi Zhu's affiliation is updated to his current affiliation

References in corpus (18)

• Unsupervised Domain Adaptation by Backpropagation
• Active Learning for Convolutional Neural Networks: A Core-Set Approach
• Cost-Effective Active Learning for Deep Image Classification
• Deep Bayesian Active Learning with Image Data
• Convolutional Neural Networks for Automated Annotation of Cellular Cryo-Electron Tomograms
• Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds
• Active Self-Paced Learning for Cost-Effective and Progressive Face Identification
• A Benchmark and Comparison of Active Learning for Logistic Regression
• Generative Adversarial Active Learning
• Discriminative Active Learning
• Improved deep learning based macromolecules structure classification from electron cryo tomograms
• Deep learning based subdivision approach for large scale macromolecules structure recovery from electron cryo tomograms
• Active Learning for Speech Recognition: the Power of Gradients
• Batch Active Learning via Coordinated Matching
• Task-Aware Variational Adversarial Active Learning
• Few shot domain adaptation for in situ macromolecule structural classification in cryo-electron tomograms
• Multi-task Learning for Macromolecule Classification, Segmentation and Coarse Structural Recovery in Cryo-Tomography
• Cost-Sensitive Active Learning for Intracranial Hemorrhage Detection

Cited by in corpus (1)

• Cryo-shift: Reducing domain shift in cryo-electron subtomograms with unsupervised domain adaptation and randomization