T2T Genome Assembly — 6.93TB dataset, in progress

We’re working along with @undiagnosed_1 on X on a full Telomere-to-Telomere assembly of a 6.93TB raw long-read sequencing dataset. The current draft has roughly 3,000 remaining gaps, and based on read depth and coverage we’re targeting Q60 consensus accuracy on the finished assembly putting it in the same tier as the reference T2T-CHM13 genome.

The pipeline were running is:

• Raw reads staged and version-controlled on GCS

• Assembly via hifiasm/verkko, ultra-long ONT reads for scaffolding across centromeric and telomeric repeat regions

• Gap-filling and consensus polishing passes to close remaining regions

• Post-assembly QC: k-mer completeness (Merqury), BUSCO gene-space completeness, and structural variant validation against known reference builds

This is a good stress test for the assembly and structure tooling and a chance to show what full end-to-end genome work looks like on top of it rather than just single-tool demos.

We’ll post updates as the assembly progresses through gap closure and QC — final assembly and analysis to follow.