*This PR is mainly about updating metabolomics-related new mets and new rxns into model.
We collected metabolomics data from 26 papers to form a metabolites pool with 1774 metabolites.
We performed name unification for those metabolites to include KEGG ID and CHEBI ID, and this step filtered away about 400 metabolites.
Based on KEGG ID and CHEBI ID, we found 845 new metabolites compared with our model. After manual check, 766 were confirmed as new metabolites.
The second filter criteria is that we only counts metabolites that appears more than twice. It ended up with 82 new metabolites.
We searched all reactions linked to those new metabolites on the MetaNetx database, and found 852 new reactions. In order to introduce least new metabolites into the model, we filtered out reactions which introduce more than 2 extra new metabolites, which ends up with 177 reactions.
Then, we chose reactions which are catalyzed by enzymes appeared in the model, only 32 reaction with EC evidence were left. Based on the criteria that we don't want to introduce dead-ends into the model, we only chose new metabolites that can be both produced and consumed and reactions linked, so it ended up with 21 reactions for 8 new metabolites.
Main improvements in this PR:
*This PR is mainly about updating metabolomics-related new mets and new rxns into model.
I hereby confirm that I have:
develas a target branch (top left drop-down menu)