Hello,
I have been using OmicsNet to integrate transcriptomics and metabolomics data and I think I may have encountered a bug regarding the expression values displayed in the network.
My workflow was as follows:
-
Genes:
#ENTREZ log2FC -
Metabolites:
#KeggID logFC(I also tried using log2FC values) -
For both datasets, I selected Metabolite–Protein (KEGG) as the interaction database.
-
I then proceeded to the visualization page.
I expected the Expression (Expr.) field of each node to display the fold change value that I uploaded for that specific gene or metabolite.
However, I noticed that the displayed expression values do not consistently match my input.
For example:
-
Some gene nodes are assigned fold change values that belong to metabolites.
-
Some metabolite nodes are assigned fold change values that belong to genes.
To further investigate, I performed another test:
-
I uploaded only a list of genes (without any fold change values).
-
I uploaded metabolites with fold change values.
Since no expression values were provided for the genes, I expected the Expr. field for every gene node to be blank (or NA).
Instead, I found that some gene nodes were still assigned expression values corresponding to metabolite fold changes.
For example:
- SAM (C00019) was assigned an expression value of −0.935325, which is actually the log2FC of the gene HOXA10 in my RNA-seq dataset.
I have also observed other instances where gene and metabolite expression values appear to be mixed.
Could you please clarify:
-
How are expression values assigned to nodes after importing gene and metabolite datasets?
-
Are expression values matched by identifier, or by the order of the uploaded lists?
-
Is this a known issue in the current version of OmicsNet?
-
Is there a recommended workflow to prevent this behavior?
Thank you very much for your help.