1
$\begingroup$

I am currently working on a network pharmacology project where I am supposed to find out the possible diseases our drug might have an effect on. I am not a bioinformatics student, so my knowledge in the subject matter is very limited. I have all the targets and I am supposed to get the pathways associated with the targets. I used STRING for the purpose.

  • I got both KEGG functional annotation and enrichment for my list of targets.
  • I have to build a two-mode network

Question I am not very clear about

  • the difference between enrichment and functional annotation and
  • which data should be used for the next step.
$\endgroup$
1
  • 1
    $\begingroup$ If you can add more detail that might help, such as the pathway(s) your STRING analysis indicates. Your question appears vague for the decision you have and it is not clear why you don't use both for two different network analyses $\endgroup$
    – M__
    Nov 22 '20 at 9:44
1
$\begingroup$

The answer to your main doubt about "annotation vs. enrichment" is that you need both, but you already have one annotation.

You can only perform a functional enrichment analysis after you have a functional annotation. In your question, you state that you have KEGG annotation for your targets. That is it, knowing which target "belongs" to each KEGG module is an example of functional annotation of your targets. Several tools can annotate genes, proteins, etc. Knowing which protein is related to each drug or disease is also a functional annotation.

Now, knowing which targets are significantly important can be achieved by functional enrichment analysis.

For the network, both information can be used together. For example, you could link all elements to their functional annotation and use the available enrichment score as a quantitative parameter of importance for calculating correlations.

$\endgroup$
-1
$\begingroup$

I would construct two networks using both data sets and compare their congruence.

  1. If there are major statistically differences between the two data sets that is the point at which you would ask you question.
  2. You then assess the biological rationale for the difference in comparison to observable differences between the networks and see whether this makes biological sense.

This is a standard statistical approach, particularly when you are uncertain of the quality of the underlying analysis in relation to your biological question.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.