Towards Foundation Models for Cryo-ET Subtomogram Analysis

2025-09-29

AI Summary

Cryo-electron tomography (cryo-ET) subtomogram classification, alignment, and averaging face three key bottlenecks: severe label scarcity, high noise levels, and poor generalizability. To address these challenges, we propose the first foundation model framework specifically designed for cryo-ET subtomogram analysis. Our method integrates: (1) CryoEngine—a large-scale synthetic data generator enabling realistic, diverse subtomogram synthesis; (2) APT-ViT—an adaptive phase-tokenized Vision Transformer that enhances structural feature representation via phase-aware tokenization; and (3) NRCL—a noise-robust contrastive learning strategy that improves discriminative capability under extreme noise and geometric/semantic variations. The framework demonstrates exceptional robustness to geometric deformations and semantic heterogeneity while achieving strong cross-dataset generalization. Evaluated on 24 synthetic and experimental datasets, it sets new state-of-the-art performance across all three tasks—significantly improving accuracy, robustness, and scalability of subtomogram analysis—and establishes a novel paradigm for high-throughput subcellular structure determination.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Deep Neural Architectures and Foundation Models
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Cryo-electron tomography (cryo-ET) enables in situ visualization of macromolecular structures, where subtomogram analysis tasks such as classification, alignment, and averaging are critical for structural determination. However, effective analysis is hindered by scarce annotations, severe noise, and poor generalization. To address these challenges, we take the first step towards foundation models for cryo-ET subtomograms. First, we introduce CryoEngine, a large-scale synthetic data generator that produces over 904k subtomograms from 452 particle classes for pretraining. Second, we design an Adaptive Phase Tokenization-enhanced Vision Transformer (APT-ViT), which incorporates adaptive phase tokenization as an equivariance-enhancing module that improves robustness to both geometric and semantic variations. Third, we introduce a Noise-Resilient Contrastive Learning (NRCL) strategy to stabilize representation learning under severe noise conditions. Evaluations across 24 synthetic and real datasets demonstrate state-of-the-art (SOTA) performance on all three major subtomogram tasks and strong generalization to unseen datasets, advancing scalable and robust subtomogram analysis in cryo-ET.

Problem

• Addressing scarce annotations and noise in cryo-ET subtomogram analysis
• Developing foundation models for classification, alignment, and averaging tasks
• Enhancing generalization and robustness for macromolecular structure determination

Innovation

• Large-scale synthetic data generation for pretraining
• Adaptive phase tokenization-enhanced vision transformer
• Noise-resilient contrastive learning for stable representation