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

Why It Matters

By eliminating the need for multiple sequencing platforms, HERRO reduces both capital and consumable costs, accelerating the adoption of T2T assemblies in research and clinical genomics. This democratizes access to complete, haplotype‑resolved genomes, enabling deeper insights into genetic variation and disease.

Key Takeaways

• HERRO raises ONT Simplex read accuracy up to 100‑fold
• T2T assemblies achieved using only corrected ONT reads
• NGA50 consistently exceeds 100 Mb across human samples
• Supports R9.4.1 and R10.4.1 flow cells
• Method generalizes to non‑human species

Pulse Analysis

The push for telomere‑to‑telomere (T2T) phased assemblies has reshaped genomic standards, but traditional workflows rely on a mix of PacBio HiFi, ONT Duplex and ultra‑long reads, inflating costs and DNA input requirements. HERRO (Haplotype‑aware ERRor cOrrection) disrupts this model by applying a deep‑learning correction pipeline to ultra‑long ONT Simplex reads, preserving haplotype‑specific signals while dramatically reducing base‑calling errors. This innovation aligns with the broader industry trend of leveraging AI to enhance raw sequencing data, offering a streamlined path to reference‑grade genomes without the expense of multiple platforms.

When paired with the Verkko assembler, HERRO‑corrected reads reconstruct up to 32 chromosomes telomere‑to‑telomere, delivering NGA50 metrics over 100 megabases—a benchmark previously attainable only with hybrid approaches. The method’s compatibility with both R9.4.1 and the newer R10.4.1 flow cells ensures flexibility for laboratories transitioning to higher‑accuracy chemistries. Moreover, the framework’s ability to generalize across species opens doors for high‑quality assemblies in agriculture, conservation, and non‑model organisms, where budget constraints often limit comprehensive sequencing.

The commercial implications are significant. By cutting the need for PacBio HiFi runs, institutions can lower per‑genome sequencing costs by an estimated 30‑40 percent, freeing resources for larger cohort studies or clinical diagnostics. HERRO also shortens project timelines, as a single ONT run suffices for both raw data generation and downstream assembly. As the genomics market continues to consolidate around long‑read technologies, tools that enhance data quality while reducing overhead will become pivotal in driving the next wave of precision medicine and functional genomics research.