search

cdanielmachado / carveme

CarveMe: genome-scale metabolic model reconstruction
Other
158 stars 53 forks source link

Reactions missing in bacterial universe after 1.5.0 #123

Closed dotPiano closed 3 years ago

dotPiano commented 3 years ago

Hi Daniel,

I could not find a changelog for v1.5.0, so this is rather a question on that.

With the latest version, some reactions that used to be in the bacterial universe disappeared (for example the RNF complex and other ferredoxin-related reactions). At first sight, it seems like a cleanup of the ferredoxin species (fdxo_42 and fdxr_42 completely disappeared). Could you clarify why this happened?

I would understand if this was necessary for consistency, just want to make sure it's not a bug :)

Best, Mattia

cdanielmachado commented 3 years ago

Hi Mattia,

Apparently, some users are reporting drastic changes in their models with the new universe: https://github.com/cdanielmachado/carveme/issues/122

I think this could be related to the same issue.

The change log is (briefly) described here: https://github.com/cdanielmachado/carveme/releases/tag/1.5.0

What is not mentioned in detail, is all the manual curation that was applied to the new universe, which might have resulted in the removal of some reactions.

The particular reaction you mention, is only present in one model from the BiGG database (iHN637). When fixing energy-generating cycles I had to remove/constrain some reactions. When in doubt, I prioritised keeping the ones that participate in multiple models. This could have been the case, but I need to check.

dotPiano commented 3 years ago

Thanks for the clarification. For now I'll work around it adding the missing reactions manually.

FYI: From what I see so far this change mainly impairs metabolism of acetogens and certain butyrate producers.

cdanielmachado commented 3 years ago

Hi Mattia,

Regarding the RNF reaction, it seems it was removed because it created an energy-generating cycle when coupled with FNRR.

This is one of the difficulties of curating the universal model, these two reactions worked fine in their original models, but cause problems when put together in the same model.