Retrieving Reactions from EC number in BiGG Model

Question

[deviinfoz.an]

Dear Biostars Team,

I have got the list of EC number for my organism 'from Enzyme Detector" (~660). So Like I need to retrieve the set of reactions corresponding to these EC number from BiGG Model (http://bigg.ucsd.edu/data_access). Why BiGG - I am working on genome scale modeling. So the model has got BiGG representation of reactions and metabolites. So once I retrieve reactions from BiGG, I need to compare it with the model and look for the presence and absence of reactions.

please help me out to sort this issue!

Thank You!

Answer 1

[nhungptt17]

Hi, I assume that you would like to map your EC numbers to reactions in BiGG reaction database to draft your genome scale model. If it is the case, then it is not really straightforward to do that. As far as I know, it does not really tell which EC numbers or which compounds participate in particular reaction when you download BiGG reaction database. I suggest to download some sbml models from BiGG and then map your EC list to EC numbers in those SBML files. Not all models from BiGG contain EC numbers for each of the reaction, but It may worth to map against saccharomyces iMM904, they have at least 600 ECnumbers that you can try.

BiGG is kind of "standard" representation of reactions and metabolites for genome scale model, but it is not always convenient to extract information from BiGG. Another less inconsistent naming system for reactions and metabolites that you can also use is KBase ID. Many genome scale models are using this format. But again, it is also not really easy to retrieve reactions from your EC list. This is a common problem in genome scale model when you want to build your model from your own annotation. You can download all reactions and metabolites from KBase here https://kbase.us/metabolic-modeling-faq/#How-can-I-download-all-the-biochemical-compounds-and-reactions-in-KBase . Depend on your organism but at least you will get the core metabolism which you can build up by adding more reactions manually.

In the worst case, you can also retrieve reactions from the most similar organism to your organism that have already modelled. Or if there is no closely related organism, there is Ecoli model.

I hope this help.

Cheers,

[deviinfoz.an]

Thanks A lot! Yeah Exactly I am in the stage of constructing a genome scale model. Actually I have got the Semi-curated version of my Organism!. So thought of refining the model again to make it organism specific by annotating the genome sequence, in specific I focus on the enzyme coding genes of my organism. So I have got around 660 Enzyme coding genes in my organism and I have got the EC numbers of the same. In this context, I thought of mapping these EC Numbers with reactions and the compare with my semi-curated model to check for the presence of reactions. If a reaction is absent, I will add those reactions in the model, sort of making it organism specific Model.

I truly agree with your suggestion like one EC Number could have more than one reactions too. I thought of working with BiGG only because most of the SBML models use BiGG model representation. I will have try the possibilities in Kbase!

Thanks Once again!

• Regards*
  • N.T. Devika*

[nhungptt17]

Hi, I just found that we can download database from modelSEED. It has more reactions (>5000 EC numbers) than KBase so maybe you can also map your EC list to it. Notice that there are some EC numbers in this list link to many reactions. You should check these reactions again. https://github.com/ModelSEED/ModelSEEDDatabase/blob/master/Aliases/Enzyme_Class.aliases https://raw.githubusercontent.com/ModelSEED/ModelSEEDDatabase/master/Biochemistry/reactions.master.tsv

Cheers,

Answer 2

[deviinfoz.an]

Hi Thanks for sharing the Reaction set from ModelSeed. But the issue is like the model which I am using don't have Model seed Identifier. So may be after mapping with Model Seed, I need to find another solution to map with the semi-curated model of mine!

Thanks Again!

[sharmatina189059]

I have list of EC numbers and I want associated reaction. For this purpose I have mapped protein ids of all ec numbers to uniprot ids of same organism. The results are like one EC number has multiple uniprot ids. Is this a correct way to retrieve reactions. How to map multiple uniprot ids to my organism's EC numbers?