Interpretation of DE gene and mutation in normal/cancer sample
2
0
Entering edit mode
9.0 years ago
bharata1803 ▴ 560

Hello all,

So after I tried to analyze DE(Differentially expressed) gene and read paper about findings of DE gene or mutation that happen in normal/cancer sample, I come up with a newbie idea (I am a newbie btw). How do we interpret the findings itself? Sometimes, we found many DE gene or mutation in normal vs cancer cell. But, actually, how do we interpret the result itself? If the findings is many (one of the paper find some mutation and also a differentially expressed gene), how can we use the information to understand more about the cancer? Also, with many biomarker can be found between normal/cancer sample, I also have a question regarding which one is actually affecting cancer or it is just affected by the cancer. Sorry if my question is so basic and simple and I thank you for your reply and answer so that it can guide me to another new knowledge.

SNP genome gene • 2.2k views
ADD COMMENT
1
Entering edit mode
9.0 years ago

There's no generic answer to your question. Perhaps the DE genes are in a few different pathways, in which case those would be interesting targets for further research or pharmacological manipulation. Perhaps the DE genes share one or two transcription factors, in which case there's probably some mutation in it/them that's driving much of the phenotype (again, that creates a drug target). There are many many many other possibilities that one could come up with. Interpreting results will always require a very good understanding of the underlying biology.

Regarding biomarkers, you're really asking about "driver" vs. "passenger" changes ("changes" to generalize the concept of driver and passenger mutations). Just start searching the literature on "driver mutations" and you should be able to generalize to other types of alterations.

ADD COMMENT
0
Entering edit mode

Thank you for your answer. The driver and passenger mutation is really helping me because I don't now many term. I will start reading about that. Also regarding gene pathway I have wondered about that too. Once again, thank you very much. I will start learn about gene pathways and also driver passenger mutation.

ADD REPLY
0
Entering edit mode

Considering the pathways of the genes, I found a research paper about transcription regulation network. Is this network represent the gene pathways? And, probably this is the another fundamental question, what are the factor that actually affect the gene transcription level? That research paper makes me think genes work is connected with each other. Do we also need to consider the relationship of gene when we analyze the DE genes besides the mutation, methylation, fusion, or expression of a single gene?

ADD REPLY
1
Entering edit mode
9.0 years ago
alolex ▴ 960

Doing a DE analysis is one of the first steps in an analysis. It gives you a global view of what has changed between normal and tumor cells. It also helps you get a global picture of the biological processes that have been affected by the tumor (or have changed in response to the presence of the tumor) through enrichment analysis. Gaining a better understanding of these global mechanistic changes is one thing to interpret from a DE analysis, but without further targeted analyses you can't move much beyond the global picture. To move beyond a large list of DE genes you have to have a specific hypothesis or question in mind that you are investigating--without a specific direction there are too many rabbit holes to run down with large lists of genes. This is where biological collaborators come in. They generally have their specific hypothesis they are investigating, so can help you narrow down the list of genes and help guide the analysis in a targeted fashion. Thus, you can interpret large DE analyses as "this is how the system has changed globally", then ask if this is what we generally expected or are there surprises we should focus on in future research.

With regards to your biomarker question. In my work, biomarkers are used to differentiate cell types, like tumor vs normal. It doesn't matter if the changes are caused by or a result from the tumor, just that they are diagnostic of the cell type (or other control vs diseased sample) so that we can make a classifier from the expression from these genes and separate unknown samples into one category or the other. There is no way to know for sure if a differentially expressed gene is caused directly by the tumor or is a side effect of the tumor's presence with prior knowledge of the system and/or further experimentation and analysis. For example, we know the body up regulates the immune system to fight off the tumor, thus, we will see these genes being up regulated in a DE analysis. However, it is not the tumor that is regulating these genes, it is the body in response to the tumor. These are the types of things that would have to be weeded out in downstream analyses and experiments for novel findings.

ADD COMMENT
0
Entering edit mode

Can you suggest a keyword or maybe aa paper in which the analysis of the "different" gene is done from the biological point of view? What I've read until now mostly is a finding about changes that occur in normal vs cancer data with Bioinformatics tools and the paper mostly talk about the method to find that and the validity of that finding. I want to understand how the Biologist analyze that information because I just can not imagine how to do that because I'm from computer science with little experience.

ADD REPLY
0
Entering edit mode

What is your biological background? I too came from the Computer Science side of things. I learned biological terminology from a biochemistry course, then the rest of it came while I was working in collaboration with biologists on specific projects. Are you working with a biologist on a project or on your own? If you have someone you are working with don't be afraid to ask lots of questions because that is how I learned most of what I know :)

For papers, I found this one doing a quick search of PubMed (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4431826/). They do a DE analysis using microarray data. Then based off of those results they narrowed in on one gene and did a knockout experiment to see how that changed things in the system. If you search on terms like "Differential Expression" or "microarray" you will get a lot of the global papers that just describe the changes. What you are looking for I think is the next step. If you do the search on Google Scholar and you find a global gene expression paper, you can look down at the links under the title for the "Cited By" link. This will pull up all papers that cite the one you found. Here you can look for the same authors to see if they published any followup work that is more biologically based--which probably would not have shown up in your initial search.

ADD REPLY
0
Entering edit mode

Well, for my Biological background, I don't have much. I take a Biology class last semester but it didn't talk too much about this kind of Biology. It's like a basic topic of Biology. I'm a master student so I'm working alone. I can talk with my supervisor about Biology thing but I can say I need to do everything myself first.

Yeah, what my supervisor think is about next step. How to interpret or to analyze the data, not just produce some list because my research is basically just from published data online in GEO Dataset which is usually the original team already cover the basic analysis step. Thank you for your suggestion. I will try your method to find similar paper like your example.

ADD REPLY

Login before adding your answer.

Traffic: 1324 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6