search

JEFworks-Lab / STdeconvolve

Reference-free cell-type deconvolution of multi-cellular spatially resolved transcriptomics data
http://jef.works/STdeconvolve/
98 stars 12 forks source link

How to make annote on my own data? #31

Closed Knight1995 closed 1 year ago

Knight1995 commented 1 year ago

it is a great tools!If i want to use my own data, i would like to know how to make the input annotation? I have seen the issue on how to annotation celltype , but i can't understand the input format of annotation and the finnal output. Actually, in my opinion, the final result should be a dataframe, for example, containing the celltype(col), cell ID(row) and the proportion. Thanks.

bmill3r commented 1 year ago

Hi @Knight1995,

Thanks so much for your interest in STdeconvolve! Could you please clarify your question? Are you asking what format your data should be in order to be used as input for STdeconvolve? Essentially the input needs to be an integer count matrix where rows are pixels (multi-cellular capture locations) and columns are selected genes for deconvolution. Does this help answer your question?

In terms of the output, STdeconvolve does return a pixel x deconvolved cell type proportion matrix, where each row is a pixel and each column is a deconvolved cell type and each cell of the matrix indicates the predicted proportion of that cell type in a given pixel. In addition, Stdeconvolve also returns a deconvolved cell type x gene expression matrix where each row is a deconvolved cell type, each column is a gene (from the selected features chosen for deconvolution), and each cell is the probability of that gene being expressed for the given cell type.

Please let me know if you have any other questions, Brendan