Any idea about projecting cell type labels back to cell segmentation plot?

[rocketeer1998]

Hi, @VPetukhov. First of all, thanks for the convenience the Baysor has brought to my ISS study. After a full baysor run, I have obtained a beautiful cell segmentation plot and a useful expression matrix that I can use in the downstream scRNA-seq analysis.

As for scRNA-seq analysis tool, I think Seurat is stabler and more robust than pagoda2, though the latter provides a convenient web app to analyze my data. So I want to perform a standard scRNA-seq pipeline using Seurat and label each cell with a cell type. So is there any recommendation about projecting this result back to the cell segmentation plot created by Baysor? Otherwise, I have to simulate a segmentation plot using ggplot2.

Thanks for your kind reply!

[VPetukhov]

Hi @rocketeer1998 , thanks for the feedback! Can you please specify, what exactly do you want to get as an output?

[rocketeer1998]

Hi @rocketeer1998 , thanks for the feedback! Can you please specify, what exactly do you want to get as an output?

Just as shown in your publication Fig.6 e and f. I want to show the spatial distribution of some cell types based on marker gene expression.

[VPetukhov]

The annotation for that was done manually, and the plot just shows each cell as a dot, so we don't have the code for that in the package. To do the same, you can take cell positions from segmentation_cell_stats.csv and match it with segmentation_counts.tsv to get cell markers.

[rocketeer1998]

The annotation for that was done manually, and the plot just shows each cell as a dot, so we don't have the code for that in the package. To do the same, you can take cell positions from segmentation_cell_stats.csv and match it with segmentation_counts.tsv to get cell markers.

OK. Thanks a lot!