Telomere-to-Telomere Assembly Using HERRO-Corrected Simplex Nanopore Reads

2026-04-27 · Source: Machine learning : nature.com subject feeds · Field: Science & Research — Life Sciences & Biology, Artificial Intelligence & Machine Learning, Mathematics & Computational Sciences · Depth: Expert, quick

Summary

The HERRO (Haplotype-aware ERRor cOrrection) framework, a deep learning-based method, enables the generation of telomere-to-telomere (T2T) phased genome assemblies using only ultra-long Oxford Nanopore Technologies (ONT) Simplex reads. Traditionally, T2T assemblies for diploid and polyploid genomes require combining high-accuracy long reads like PacBio HiFi or ONT Duplex with ultra-long ONT Simplex reads, increasing cost and DNA requirements. HERRO corrects ONT Simplex reads while preserving haplotype differences, achieving up to a 100-fold increase in read accuracy for diploid human genomes. When combined with the Verkko assembler, HERRO reconstructs up to 32 chromosomes telomere-to-telomere, including X and Y, and consistently yields NGA50 values of 100 Mb or higher across multiple human genomes. The framework supports both R9.4.1 and R10.4.1 ONT Simplex reads and is generalizable to other species, demonstrating a path to lower sequencing costs and improved genomic analysis quality.

Key takeaway

For genomics researchers aiming to achieve high-quality telomere-to-telomere genome assemblies, HERRO offers a significant advancement. You can now achieve benchmark-quality results using only ultra-long ONT Simplex reads, potentially reducing the need for multiple sequencing platforms and lowering overall project costs. Consider integrating HERRO into your assembly pipeline to improve read accuracy and achieve more complete, phased genome reconstructions, especially for complex diploid or polyploid genomes.

Key insights

HERRO enables high-quality telomere-to-telomere genome assembly using only error-corrected ONT Simplex reads.

Principles

• Haplotype-aware error correction improves assembly quality.
• Single-platform sequencing can achieve T2T assembly benchmarks.

Method

HERRO uses deep learning to correct ultra-long ONT Simplex reads, preserving haplotype differences, then combines these corrected reads with a de novo assembler like Verkko for T2T assembly.

In practice

• Use HERRO for cost-effective T2T genome assembly.
• Apply HERRO to R9.4.1 and R10.4.1 ONT Simplex data.

Topics

• Telomere-to-Telomere Assembly
• ONT Simplex Reads
• HERRO Framework
• Deep Learning
• Haplotype-aware Correction

Best for: AI Scientist, Research Scientist

Related on AIssential

• Large genome model: Open source AI trained on trillions of bases — Evo 2 is an open-source AI trained on trillions of DNA bases to identify complex genomic features across all life domains.
• Prediction and functional interpretation of inter-chromosomal genome architecture from DNA sequence with TwinC — TwinC predicts inter-chromosomal DNA contacts, revealing their mechanistic drivers and role in gene regulation.
• Streaming Speech-to-Text Translation with a SpeechLLM — A novel SpeechLLM adaptively decides when to emit translation tokens, achieving low latency and high quality.
• Probabilistic modelling of single-cell bisulfite sequencing data with MethylVI — MethylVI disentangles biological signals from technical noise in single-cell bisulfite sequencing data.
• Rethinking Genomic Modeling Through Optical Character Recognition — OpticalDNA reframes genomic modeling as OCR-style document understanding, achieving efficiency and high-fidelity compression via visual layouts.
• Improved tumor-only variant calling and mutation burden estimation with VarNet-T — VarNet-T improves tumor-only somatic variant calling and TMB estimation using weakly supervised deep learning.

Open in AIssential →

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