Deconvolution of cancer subtypes

[ktrannt]

Hi guys,

I am using BayesPrim to deconvolve a breast cancer dataset, in which there are 5 sub-types of cancer epithelial cells.

I'm more interested in the cancer sub-type populations (cell.states) than the cancer population itself (cell.types).

Just wondering if there's any way I can get the updated cell.states fractions? Or perhaps an option to specific multiple keys for cancer populations in new.prism()?

Cheers, Khoa.

[tinyi]

Hi Khoa,

Thank you for your interest in our work.

BayesPrism assumes that the malignant expression in each tumor sample is different, by summarizing a single profile of malignant cell in each bulk when performing the update and hence the cell state information of the malignant cells is lost. One workaround is that you may use the initial estimates of the cell states to approximate the cell state fraction.

Best,

Tinyi

On Wed, Jul 27, 2022 at 11:28 PM Khoa Tran @.***> wrote:

Hi guys,

I am using BayesPrim to deconvolve a breast cancer dataset, in which there are 5 sub-types of cancer epithelial cells.

I'm more interested in the cancer sub-type populations (cell.states) than the cancer population itself (cell.types).

Just wondering if there's any way I can get the updated cell.states fractions? Or perhaps an option to specific multiple keys for cancer populations in new.prism()?

Cheers, Khoa.

— Reply to this email directly, view it on GitHub https://github.com/Danko-Lab/BayesPrism/issues/8, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB4NHSYEBTIAT3DGZX327QLVWH46JANCNFSM543W54LQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>