WGCNA lncRNA-mrna interaction

Question

[Azhar]

I am new to R want to know that I have already normalized data for lncRNA and mrna means Excel files for different regions for Male and female I want to first find out lncrna mrna interaction and build network and then compare 1.I want to find out pearson correlation for each pair of lncRNA and mRNA (Note:Two lists are not equal) How can I find out pair wise corealtion ??? I have to compare each each and every mRNA against each of lncRNA? Second point is do I need their attribute to compare like FC or Exp Value or just name list for both of them is enough? ;

[GouthamAtla]

What do you mean by "lncRNA and mrna means Excel"

[Azhar]

I mean normalized data differential expressed lncRNA list and mRNA list with their Fold change and P value

Answer 1

[GouthamAtla]

Your title has answer.

Use WGCNA to find the co-expressed lnRNAs and mRNAs. If you have enough sample size ( min. 20 samples recommended ), you could run WGCNA and identify the tightly co-expressed LncRNAs and mRNAs.

WGCNA had pretty detailed documentation and several lectures on youtube from the author.

[Azhar]

I have normalized data and how to use WGCNA FOR MY DATA AS SATATED ABOVE ? WHY WE NEED 20 SAMPLES ? SECOND IS i AM USING CYTOSCAP GENE EXPRESSION COREALTION BUT WHEN I BULD NETWROK FOR ONE MNRA WITH ITS ATTRIBUTE pVALUE AND LNCrna PVALUE IT SAYS YOU HAVE VALUES LESS THEN 4 WILL PRODUCE ERROR

WHY?

[GouthamAtla]

You need 20 samples for statistical significance. If you have a matrix, 15 - 20 samples ( columns ) and mRNA/LncRNA expression values ( rows ) , you could follow the WGCNA tutorial.

[Azhar]

if i have 8 samples of mrna of one class and 8 samples for lncRNA , so i can i use WGCNA ?? and what value for each samples i should choose Pvalue ? and then i want to build network and compare for signficant difference between two groups

[Azhar]

How can I follow which tutrioal can you tell me please because I see this link and read tutorial and 1st is

I. Network analysis of liver expression data from female mice: finding modules related to body weight

but this uses only mrna expression data I need to find out lnc rna vs mrna paired Pearson correlation and build network and then compare these networks

[GouthamAtla]

If the data that you have of mRNA and LncRNA are of same samples, you could simply combine them and follow tutorial. In general, when you do RNA-Seq, you could quantify both mRNAs and LncRNAs.

Load the raw counts in DESeq2 and get the VST data and feed this vst data to WGCNA. Then it will detect modules of co-expressed gene, and check if the LncRNAs and enriched in any module.

[Azhar]

Thanks Bro Really Helpful can you guide me more ! yes I have same sample means coulmn names for lncRNA and and mRNA like

mRNA gg aa cc kk lncRNA gg aa cc kk 6.7 5.8.NA 9.7 8.9 5.6 9.9 NA you mean i can combine them in one file and use

I. Network analysis of liver expression data from female mice: finding modules related to body weight

method for WGCNA right then i can comare two gropus (male and Female ) by 2. Consensus analysis of female and male liver expression data from WGCNA

and second method is different right ? i dont have raw data file i only have normalized data files so DESeq2 can be used for this data also or not? I have raw expression values for already annotated diff expressed genes and lncRNA is this data can be used as VST data