Differential gene co-expression analysis using scLink

[ismailelshimy]

Dear @Vivianstats ,
Could you kindly explain briefly how scLink can be used to perform differential co-expression analysis between case and control samples similar to what you performed using the breast cancer data in your manuscript?

Did you previously try to construct one single network from all gene expression data from both cases and controls and then create two separate networks one for cases and one for controls and compare each of these separate networks to the combined network using all samples?

I would really appreciate if you could clarify your recommended approach for this type of analysis?

[Vivianstats]

Hello,

We didn't try to construct one single network from all gene expression data.

Please also note that, in our real data study on the breast cancer, we assessed the statistical significance of the scLink correlation using a permutation approach, but we didn't assess the significance of the differential co-expression. For any correlation measure, a permutation approach can be designed to assess the significance of differential co-expression, but it might be a time-consuming approach is many cases. There are also a few other methods designed to perform differential co-expression analysis, but we haven't tried them in the scLink study, so I cannot comment on their effectiveness.

[ismailelshimy]

@Vivianstats sorry for my late reply, I would like to thank you very much for your clear explanation. It is greatly appreciated.