Understanding the various outputs of the inferCNV HMM

[Daniel]

Hello and thanks for the great tool you have put together!

I've gone through the InferCNV workflow and added the HMM output to my Seurat object's metadata. The tutorial for doing this (https://github.com/broadinstitute/inferCNV/wiki/Extracting-features) mentions two types of data produced by the HMM calls: proportion_scaled_(loss/gain)_(chromosome number), and top_(loss/gain)_(chromosome number).

By inspecting the metadata of my Seurat object, I see these types of features as well as has_cnv_(chromosome number), has_(loss/gain)_(chromosome number), and proportion_scaled_(chromosome number). Is there anywhere I could get more information as to what all these features mean and how they are generated?

Additionally, I was wondering if there was any way to cross-reference the CNV region prediction reports mentioned here (https://github.com/broadinstitute/inferCNV/wiki/inferCNV-HMM-based-CNV-Prediction-Methods) when adding metadata to a Seurat object. I am interested in arm-level or sub-arm-level CNV events, and I can't seem to get any information other than if a gain or a loss is present somewhere on the chromosome from the current tutorial. Ideally, I would like to be able to pull out something like has_loss_1p from my Seurat metadata.

Thanks in advance!