I am currently doing a project investigating the lipid accumulation potential of a chemical using an in vitro model.
Upon observing significant lipid droplet accumulation at various concentrations, a concentration was chosen to undergo RNA-seq analysis on the cells at various timepoints of exposure.
Principal component analysis showed a clear distinction between the treated and negative control at the later timepoint. Statistical analysis has been carried out and it all seems ok.
However, no significant pathways have been identified by a gene ontology analysis that could explain the increase in lipid so I was wondering if it would be possible if there is a mechanism for the lipid homeostatic balance to be disturbed not through direct gene expression changes but by another means such as proteomic for example? I am not an expert in pathways relating to this field and so I was wondering if someone had ideas?
Hello Raquel.
It would be easy to conclude that non-genetic regulation may be present. I would say there is no such regulation except for degrading by ROS through the membranes from the medium (containing everything in the surroundings) or generated by the cells.
It would not be enough to get an insight into real lipid metabolism by looking at the predicted statistical pathways that are based only on the knowledge we have constructed so far.
It is more reasonable to think a specific pathway of a certain form of lipid molecule alone each time. The fact that total amount of lipid accumulated in lipid droplets or vacuoles or some other organelles within cells is a sum of activities of many biosyntheses, lipidases, desaturases, saturases, lipolysis, etc, not to mention huge carbohydrate metabolic enzymes. Therefore, transcriptome analysis that looks at 10~30k genes together would not match well to total lipid analysis. Individual genetic regulation would be completely masked in the total lipids. We simply have no knowledge of real parts of inter-regulatory relationships among the pathways, among the intermediate metabolites, and any positive or negative feedback mechanisms within a single pathway unlike the predicted genetic pathways. In body, there are more tangled factors, collectively called 'physiology'.
My suggestion is just to look at single lipid molecule at a time separately in your analyses to find which molecules are most affected in top 10 genes or bottom 10 genes related to lipid metabolism to try to find possible inter-regulatory mechanism(s).
Best wishes.
Sang Ho Lee, thank you very much for taking the time to respond especially in such a thorough manner. I sincerely appreciate it and your advice. I will definitely take time to break down the complexity and investigate lipid individually and see if there is any relevant literature available on what I find.
It is my first time analyzing raw data like this so I am getting to grips with trying to understand the various factors to consider both in being able to analyse as well as limitations of both software and available knowledge.
Thanks again,
Best wishes
:)
Hi Raquel,
To try to provide a little help, I will need to ask you a couple of questions
First, how did you do the gene ontology analysis as depending on the method not all of them have the same power to identify the dysregulated GOs.
And second, which input data you used to carry on the Gene ontology analysis did you selected the most dysregulated genes and carried on a Singular Enrichment Analysis (SEA), or you used all the RNASeq data and carried on a Gene Set Enrichment Analysis (GSEA) or did a and “Modular Enrichment Analysis (MEA)?
Depending on your answers there could be different analytical options. But if you think that no other methodology to re analyse the association between your RNASeq data and the GO terms will be successful then maybe you can try to do:
1. As Sang Ho Leesuggested identifying if your top dysregulated genes are associated in any specific lipid related Go term.
2. Use a Protein protein interaction databases as STRING to identify interactors of these top genes and again assign them the GO terms and see if any is involved in lipids metabolism.
3. Identify all the genes associated with GO terms involved in lipids, find them in your RNASeq data and assess their expression and regulation
Best,
Mar
Dear Mar,
Thank you for taking the taking to get back to me. Your advice is extremely useful! :)
Ended up doing GSEA; a pattern emerged in our data. The gene ontology analysis highlighted pathways which could explain the increases in lipid which has encouraged some ideas for future experiments to validate if the theme identified.
Identification of the genes with the GO terms and further analysis will also been done :)
Best,
Raquel
- Badges
- Science topic
- Similar topics
- Biological Science
- Genetics
- Gene Expression