EMReady2: improvement of cryo-EM and cryo-ET maps by local quality-aware deep learning with Mamba

2026-05-16 · Source: Machine learning : nature.com subject feeds · Field: Science & Research — Life Sciences & Biology, Artificial Intelligence & Machine Learning, Mathematics & Computational Sciences · Depth: Expert, short

Summary

EMReady2 is a new deep learning method designed to enhance the quality and interpretability of cryo-electron microscopy (cryo-EM) and cryo-electron tomography (cryo-ET) maps. This extension of the previous EMReady method incorporates a fast Mamba-based dual-branch UNet architecture, enabling it to capture both local and global features within the maps. A key innovation is its local resolution-guided learning strategy, which specifically addresses the inherent quality heterogeneity found in cryo-EM maps. The training dataset for EMReady2 has been significantly expanded, making it applicable to a wider array of maps, including those containing nucleic acids and medium-resolution maps. Evaluated against 136 diverse maps ranging from 2.0 to 10.0 Å resolutions, EMReady2 demonstrates superior performance in improving map quality and interpretability while also reducing computational costs compared to existing post-processing techniques.

Key takeaway

For structural biologists and cryo-EM/ET practitioners seeking to improve map quality and interpretability, EMReady2 offers a computationally efficient solution. Its Mamba-based architecture and local resolution-guided learning strategy can significantly enhance diverse maps, including those with nucleic acids or at medium resolutions. Consider integrating EMReady2 into your post-processing workflow to achieve clearer structural models and reduce analysis time.

Key insights

EMReady2 enhances cryo-EM/ET map quality and interpretability using a Mamba-based UNet and local resolution-guided learning.

Principles

• Address map heterogeneity with local quality awareness.
• Combine local and global feature capture for robust improvement.

Method

EMReady2 uses a Mamba-based dual-branch UNet architecture with a local resolution-guided learning strategy and an expanded training set to improve cryo-EM/ET map quality.

In practice

• Apply EMReady2 to improve medium-resolution cryo-EM maps.
• Utilize EMReady2 for cryo-ET maps containing nucleic acids.

Topics

• EMReady2
• Cryo-EM Map Improvement
• Cryo-ET
• Deep Learning
• Mamba Architecture

Best for: AI Scientist, Research Scientist, Machine Learning Engineer

Related on AIssential

• CryoACE: An Atom-centric Framework for Accurate and Automated Model Building in Cryo-EM — CryoACE offers an atom-centric, training-free approach for accurate and automated protein modeling from cryo-EM data.
• Orthrus: toward evolutionary and functional RNA foundation models — Orthrus is a Mamba-based RNA foundation model using contrastive learning and biological augmentations for superior RNA property prediction.
• Latent space mapping of interpretable structural coordinates from stochastic single-molecule signals — Latent-space mapping resolves stochastic signal warping in nanopore sensors, enabling interpretable molecular coordinate systems.
• Author Correction: Mask-prior-guided denoising diffusion improves inverse protein folding — A published correction clarifies MapDiff's amino acid prediction accuracy in inverse protein folding.
• 🔬ESM: The Bitter Lesson is Coming for Proteins - Alex Rives, BioHub — Scaling protein language models with vast, diverse data reveals emergent biological capabilities and enables programmable design.
• Augmenting Molecular Language Models with Local $n$-gram Memory — Integrating local $n$-gram memory via MolGram efficiently resolves the locality gap in molecular language models, improving performance with fewer parameters.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine learning : nature.com subject feeds.