How to analysis metabolomics data with pathway? Why library doesn't require my dataset?

Question

[chanwoo.park]

Hi. I'm analyzing metabolomics data concerned with liver cancer data.

I'm wondering how to analyze this data. (I want to get p-value of pathway that I already setted)

More specifically, why MetaboAnalyst, FELLA, such like these library do not require my dataset?

I think that if we want to find significant pathway, we must take care about which person having liver cancer.

My dataset form is like this.

Firstly, I have each sample and metabolites value, and whether or not having liver cancer. Second, I have the list that which metabolite is concerned in which pathway.

Thank you.

[ATpoint]

In StackExchange language one would say This question needs details and clarity. Unclear to me what you are asking.

More specifically, why MetaboAnalyst, FELLA, such like these library do not require my dataset?

What does that mean?

[chanwoo.park]

if we put metabolite name, then it gave us pathway's p-value. But I think this does not consider about my dataset.

[Egon Willighagen]

What does your data look like? Do you have metabolite identifiers? From which database?

Answer 1

[Elucidata]

In MetaboAnalyst, pathway analysis can be done in two ways:

• You can provide a list of metabolites (KEGG IDs, HMDB IDs, or compound names). It does an over-representation analysis that tests whether compounds involved in a certain pathway is enriched compared by random hits. These metabolites could be differentially expressed metabolites for the cohort comparison under investigation (in your case liver cancer vs normal).
• You can provide a data frame with metabolites in columns (or rows) and samples in rows (or columns) along with the cohort info that describes the condition individual sample belongs to. It would do a quantitative enrichment analysis directly based on the concentration values of the compounds.

You can refer to the FAQs that talk in detail about the functionalities.