I used EdgeR to get the list of DEGs (differentially expressed genes) from my liver RNA-seq experiment. I have done the gene overrepresentation and GSEA based pathway analysis. I am currently taking one gene at a time and googling if someone has shown its role for something similar to my research topic. My question is, how do I filter important genes from unimportant ones (like olfactory receptors, immunology related, and predicted genes)? How do you guys do it? Please give valuable suggestions.
Hi Ankur B. Sharma , I do not think there is correct answer to your question, since terms like "important" or "unimportant" require a biological context. From your description, it sounds like you have identified some "important" pathways (or, gene-sets), or those that are disproportionately represented in your list of differentially-expressed genes. While "olfactory receptors" may not directly pertain to your analysis, the genes responsible for driving this enrichment may be relevant to a parallel pathway in another tissue-type. Thus, depending on the pathway database, you can identify the genes responsible for driving this enrichment. To do this, Enrichr is convenient (both the web-based and R platforms), exporting the pathway as a table (.txt).
Another approach is to search for "nodal" regulators of gene expression. Common tools include transcription factor binding site (TFBS) analysis, motif analysis, and even weighted gene coexpression network analysis (WGCNA). To accomplish this, you can use TRANSFAC, JASPAR, ENCODE, ChEA, etc...
- Badges
- Science topic
- Similar topics
- Control Systems Engineering
- Filtering