Why is there no curated genome-wide array CNV callset (comparable to GATK-SV) in All of Us?

Hi All of Us genomics team / community,

I’m working with structural variant / CNV data for phenotype association analyses and had a question about the current CNV landscape.

All of Us currently provides:

• A high-quality, cohort-level srWGS SV/CNV callset via GATK-SV (~96–98k participants; ~1.5M SV sites)
• lrWGS SV calls (PBSV / Sniffles2) on a smaller subset
• Genotyping array data at much larger scale (hundreds of thousands of participants), including intensity information (e.g., LRR/BAF) and intensity-only probes

What I don’t see is a curated, genome-wide array CNV VCF that is jointly called / QC’d at a level comparable to the GATK-SV short-read SV release (i.e., sample-level genotypes for deletions/duplications, site-level filters, and cohort allele frequencies).

Questions:

1. Is there a technical or design reason All of Us has not released a curated array CNV callset?
2. Is array CNV calling planned for a future CDR release (e.g., as a companion to GATK-SV), or considered out of scope?
3. If researchers want to call CNVs from the GDA / IDAT intensities themselves, is there recommended guidance, recommended software, or known caveats for All of Us array data?

Why this would help: An array CNV resource would substantially increase CNV sample size beyond the current ~97k GATK-SV set (especially valuable because CDRv9 expanded srWGS SNP/indel data far more than SV coverage). Even if arrays only reliably capture larger CNVs, a curated callset would support discovery, replication, and rare large CNV analyses linked to EHR phenotypes.

Thanks in advance for any clarification on roadmap or rationale.

Best,

Shicheng