Interpretation of Kanssandra output

[WuRAFY]

Hi, Firstly, thanks for the tool. Recently, I tried some deconvolution tools like BayesPrism, SCDC and MuSiC on LUAD. The results of these tools show a quite low lymphocyte percentage (<1%). According to explanantion on BayesPrism github page, "BayesPrism infers the reads from each cell type". As lymphocyte expresses way lower amount of reads, the lymphocyte percentage inferred by BayesPrism could be quite low.

I'm glad to see a more reasonable lymphocyte percentage in Kanssandra ouput. But I still wonder the meaning of Kassandra output is A or B:

e.x.  50% T cell + 50% malignant cell
A 50% reads from T cells and 50% reads from malignant cells  
B 50% cells are T cells and 50% cells are malignant cells

Thanks, F.Y.