Closed ptdtan closed 1 year ago
Hi @ptdtan, sure, I'll upload the code for that.
Just an update on this. Now I can reproduce your result in Fig3 for sample Lung13. My issue was that I tried to use the coef
vector of each cell type in the DIALOGUE result to compute the product of the gene expression vector of each cell in the corresponding cell types. I intended to use those products as the MCP scores.
But I found that the DIALOGUE result already has the scores
for each MCP in each cell type. I used that to plot and viola.
But I'm still looking forward to double-check with your reproducible scripts :D
Attached are my plots. I only use the expression matrix of Lung 13 and the sample identity is the fov
MCP1 on Lung13
MCP1 negative on Lung13
OK, great - glad you were able to reproduce the results. Thanks for posting!
Hi @livnatje, We are trying to reproduce the Nanostring SMI results in the Nature Biotechnology paper but seems that our results are quite different compared to what was reported in the paper. The sample is Lung 13. Do you have the reproducible code to produce Figure 3a, 3b in the paper.
Many thanks for the great method.
Tan.