How could I get the ground truth of GSE111672?

jianhuupenn / SpaGCN

SpaGCN: Integrating gene expression, spatial location and histology to identify spatial domains and spatially variable genes by graph convolutional network

MIT License

199 stars 60 forks source link

How could I get the ground truth of GSE111672? #42

Closed amssljc closed 2 years ago

amssljc commented 2 years ago

Thanks for your work! How could I get the ground truth of GSE111672? It is a Primary pancreatic cancer tissue dataset, but I can only get the expression matrix. Also, I am a fresh in ST area. Could you tell me how to assign coordinates such as '10x30' in the expression matrix to the spatial image which has 30000x40000 pixels?

joungmin-choi commented 2 years ago

Hi Dr.Hu, Thanks for your great research. I also have a same question with the above. Where can we get the ground truth (pathologist-annotated cancer region) information for GSE111672? It would be really helpful if you could please let us know.

jianhuupenn commented 2 years ago

Thanks for your interest in SpaGCN. The annotation is from the figures in the original publication. In this case, you need to do the manual mapping on your own to transform the array coordinates into pixel coordinates. I didn't find any existing package that can help you with this.