Hi,

You can focus on logFC and p-adj value treshold. There is no standard values for that, it depends on your data set, the number of DE genes, the level of expression of your genes (genes with low expression can have higher logFC for example). You could try to refine your lists of genes considering genes with logFC > 2 or < -2 for example. Or narrow down your list with a p-adj < 0.01 or 0.001 (meaning that the false discovery rate would be < 0.01 or 0.001).

With your list of DE genes, you can also go for an enrichment analysis. The GO consortium provides a tool for that although not all sequenced species are implemented yet: http://www.geneontology.org/

You could also look at the GO terms associated with your DE genes..

These are just first suggestions, there is quite a lot you can do to fully analyze a RNA-seq dataset.

Best,

Hi @ Audrey Mat

Actually I have done the GO and KEGG enrichment with DEGs already. But I have seen some papers in which they used heat map to analyze the gene expression profile changes along the stressed time duration. 

Yes, you can also go with that. Or cluster analysis.

Hi @Audrey,

That is what I want to do. Do you have any good papers or tools to recommend for this kind of analysis?

Hi,

Check the DESeq2 manual, there is a section on that, section 2.2 Data quality assessment by sample clustering and visualization. There is also heatmap.2 from the gplots package, but I have not personnaly used it. As for RNA-seq analysis in general, there is no standard workflow or method, but you might find more help on the bioconductor website and other forum for detailed codes.

thank you @ Audrey.