simultaneous co-expression network for various cell-types

[AkilaRanjith]

Thanks for intuitive package for co-expression network analysis.

I have some doubts in setting the co-expression network i would like to generate co-expression for 8 major cell types, So can i give all the major cell types in 'group_name' parameter or i have to loop the code for each cell type I am using HPC so memory wont be an issue

Thanks Akila

[smorabit]

Hi Akila, this is a good question. The SetDatExpr function takes the group_name parameter to select the cell population(s) of interest for network analysis. If you pass more than one cell type to this parameter, it will make a co-expression network using cells from both of these groups. Since you want to make a different co-expression network for each of your 8 major cell types, you are going to want to write a loop or run in parallel as separate jobs since you're on an HPC.

In this script for the hdWGCNA paper Figure 1 you can see how we just ran a loop to make a different network for each cell type. You may also wish to subset your Seurat object to just contain your cell type of interest before running SetupForWGCNA.

[AkilaRanjith]

Hello Sam,

Thanks for developing the intuitive co-expression network “hdWGCNA” . I have read your bioRx version.

Following to my Github question about simultaneous co-expression network for various cell-types https://github.com/smorabit/hdWGCNA/issues/82#top

82

I have some more queries regarding the module generation. I Would like to generate two types of module networks

1. Network between 6 major cell types (inter-module interaction between major celltypes) for this can I use setDatExp like the following syntax

seurat_obj <- SetDatExpr(seurat_obj,group_name = c("Inhibitory","Excitatory","Microglia","OPC","Astrocytes",” Oligodendrocytes”),group.by='major_celltypes', slot = 'data',wgcna_name="major")

2. Generate a network based on granular cell types within each major cell type (similar to Figure 4) in the bioRx paper. (Within-Module interaction)

   For this can I loop each major cell type subset?

For obtaining hub genes the program uses degree or it also considers betweenness centrality measures? While plotting the hub gene labels in the network become overcrowded can you use ggrepel or ggraph to improve the plotting?

I know this tool is still under development and you are working on the tutorial,but I would like to explore the functionality. Please share your thoughts!!!

Thanks Akila

On Sat, Feb 11, 2023 at 9:51 AM Sam Morabito @.***> wrote:

Closed #82 https://github.com/smorabit/hdWGCNA/issues/82 as completed.

— Reply to this email directly, view it on GitHub https://github.com/smorabit/hdWGCNA/issues/82#event-8497053005, or unsubscribe https://github.com/notifications/unsubscribe-auth/AF6ZH3JNNQ7234UQYKPGVA3WW7GT3ANCNFSM6AAAAAAUV452PU . You are receiving this because you authored the thread.Message ID: @.***>

-- Thanks with Regards S.Akila Parvathy Dharshini, Post-doctoral Research Fellow, Department of Pathology, Stanford School of Medicine