search

Vivianstats / scLink

Inferring gene co-expression networks from single cell gene expression data
https://www.sciencedirect.com/science/article/pii/S1672022921001455
24 stars 7 forks source link

Retrieve gene modules #2

Open gouthamatla opened 3 years ago

gouthamatla commented 3 years ago

Hi,

I am wondering if there are any functions to retrieve gene modules from network net1$adj or any plotting functions.

Thanks, Goutham A

Vivianstats commented 3 years ago

Hi Goutham,

The scLink package itself does not contain functions for module detection or plotting, but you can use the igraph package to perform downstream analysis after you obtain results with scLink.

-Vivian

gouthamatla commented 3 years ago

Hi Vivian, thanks for the quick reply.

I did run scLink on a scRNA-Seq data. The resulting network is very large and uninterpretable. I am wondering if there are any recommendations to split or get only high confidence edges from the network. I understood that one thing is to look for correlation among genes and use genes that are highly correlated. Is that correct ? I am not an expert in network biology. Any pointers would be really helpful.

I guess my question is about community detection. Should I use the concentration matrix as a weighted matrix for community detection ?

Vivianstats commented 3 years ago

For community detection, you can either use the correlation matrix given by sclink_cor (then every two genes will have an edge with the weight being the robust correlation calculated by scLink) or the sparse network (in the $adj entry) given by sclink_net (then the edges are sparse and not weighted).