Incomplete assembly in HPRC dataset

Question

[ohell]

Hello,

We were investigating the mapping of some sequencing data simulated from known sites across populations. We decided to use the HPRC pangenome (the minigraph cactus one, v1.1). Saw some multi-mappings that did not make sense, e.g. missing ALAS1 gene in South Asian population.

Investigation led me to discover that the single South Asian assembly included in the HPRC pangeome v1 is missing 6 chromosomes including chr3 where ALAS1 is located! We're wondering if omissions like this are documented somewhere, and whether v2 dataset also includes incomplete assemblies?

This somewhat damages our starting assumption that pangenome haplotypes reflect the occurrence in populations, that is all :)