Infer malignant cell fraction using CNV, but infer normal cell fraction with matched scRNAseq dataset?

[wlwang20]

For the deconvolution of my ST data, I want to use my own scRNAseq reference for the normal cell cell fraction. However, I am satisfied with the way that the default SpaCET.deconvolution is determining the fraction of malignant cells using the CNAs. Is there a way to blend the two methods (CNA for malignant fraction deconvolution and matched scRNAseq for the normal cell fraction deconvolution)?

It seems like SpaCET.deconvolution.matched.scRNAseq does both malignant and normal cell deconvolution, but how can I make it work only on the normal cell fraction? Thanks.

[beibeiru]

Hi, @wlwang20,

Thank you very much for your interest in our tool!
I design two new parameters sc_includeMalignant and cancerType for SpaCET.deconvolution.matched.scRNAseq, which can enable to infer malignant cells using CNA dictionary and non-malignant cells using the input scRNAseq reference. Just set sc_includeMalignant as FALSE and cancerType as the cancer type of your data.
Please reinstall our package and have a try. Thanks.

Best,
Beibei

[wlwang20]

Thank you! This update worked for me!