NA genome assembly for some genomic mutations

[Matteo Tiberti]

Dear cBioPortal and users,

I am retrieving POLD1 mutations through the cBioPortal API and found a few mutation occurrences for which the detailed response reports ncbiBuild as NA.

Four occur in msk_impact_50k_2026_mutations and three in ucec_ccr_msk_2022_mutations. They are:

A171V chr19:50905304C>T
K648N chr19:50912430G>T
T1017R chr19:50919963C>G

The coordinates are consistent with the GRCh37/hg19 POLD1 locus.

In the official msk_impact_50k_2026 DataHub mutation file, the POLD1 rows specify NCBI_Build=GRCh37, while the API returns NA for the affected occurrences. 

In the ucec_ccr_msk_2022 mutation file, the build is blank for the affected records and appears blank throughout the POLD1 subset. 

I found these issues just working on POLD1, but I'll have to repeat the same data fetch for many other (hundreds) of genes so this is more a general problem for me.

Would it be safe to default all the entries I find as NA to having been built on GRCh37? I suspect these come before a time when hg38-defined mutations were available, but I'm not sure. Any other way around this?

Thank you in advance

_____________________________

Matteo Tiberti, PhD

Staff Scientist at the Cancer Structural Biology group
Danish Cancer Society Research Center

Strandboulevarden 49

2100, Copenhagen, Denmark