Networks

Networks for RNASeq data

The Rscript "coexpression_Spearman.R" can be used to create a coexpression network based on gene expression.

Methodology:

1. Calculates Spearman correlation for each pair of genes
2. Save the results in a large matrix
3. You need to choose a threshold in order to get the most co-expressed genes (i.e: > 80 of Spearman)
4. Save the results in a Rdata file, then you don't need to run everything again
5. Generate heatmaps
6. Create the final file with the gene source, gene target and weights.

Now, you can open it on Cytoscape and have fun!

We've have used here an example of expression data from different tissues of chicken (E-MTAB-2797: public avaible at ArrayExpress). However, you can use your own expression data. Just make sure that you have filtered the non-expressed genes and have normalized the data.

You can see a general figure here and a zoom here!

---

Created by:

• Katia de Paiva Lopes
• Bioinformatician, PhD