Discrepancy between OmicsAbsoluteCNGene and OmicsAbsoluteCNSegmentsProfile

Hi,

I was looking at OmicsAbsoluteCNGene values (24Q4) for the gene DCC on chromosome 18q, in relation to the segment value in OmicsAbsoluteCNSegmentsProfile (also 24Q4). For cell line ACH‑000679 (profile ID PR-pjbPYg), I am finding that the OmicsAbsoluteCNGene = 1, but if you map the gene coordinates for DCC into the segments file, OmicsAbsoluteCNSegmentsProfile would predict that the segment to which DCC maps onto has a SegmentAbsoluteCN of 2 and an LoHStatus of being copy-neutral LOH.

The genomic coordinates for DCC are chr18:52340197-53535899 (hg38) or chr18:49866567-51062269 (hg19). I’m not sure what mapping was used to generate the OmicsAbsoluteCNGene values, but both the hg19 and hg38 coordinates map to a region of SegmentAbsoluteCN = 2 in the segment file. My first thought was that the discrepancy was just due to me using different genomic coordinates than what was used in the original pipeline. However, I’m not sure the exact genomic coordinates that were used in the original pipeline are available as a source file (please direct me there if that is available!).

Any help on understanding how the OmicsAbsoluteCNGene calls were made would be helpful. For context, for my research project it would be very important to distinguish between copy-neutral LOH and copy-deletion LOH.

Hi,

OmicsAbsoluteCNGene is indexed by ModelIDs while OmicsAbsoluteCNSegmentsProfile is indexed by ProfileIDs. When a model has both WGS and WES profiles, usually the WGS is selected as the default to represent the ModelID. In the case of ACH‑000679, its WGS profile is PR-bDxA3a and its WES profile is PR-pjbPYg, and the default WGS profile is represented in OmicsAbsoluteCNGene. It appears that the segment-level CN for PR-bDxA3a does agree with OmicsAbsoluteCNGene. Due to the nature of different sequencing methods and the time and place of each sequencing, we do expect some level of disagreements between different omics profiles.

Our copy number uses hg38 as the reference genome so your coordinates shouldn’t be a problem here. Let me know if this clears things up.

Thanks,

Simone

Thank you, that clarifies things!