carveme VS embl_gems

Carveme: https://academic.oup.com/nar/article/46/15/7542/5042022 https://github.com/cdanielmachado/carveme https://carveme.readthedocs.io/en/latest/usage.html

Thanks sharing, I believe I have seen the github repo/paper before! Looks very interesting, I wonder if mantis could be used for expanding the functional annotation of sequences in the BiGG database (paper). This is the database that is used by CarveMe (paper) for building metabolic models from protein fasta files within the metaGEM workflow. I believe that CarveMe does not actually use the annotations themselves, but simply aligns sequences in a fasta file to the BiGG database, which contains associations between sequences -> genes -> reactions -> metabolites.

Hi OP, what is your organism? Is it possible that a model already exists for that organism here (https://github.com/cdanielmachado/embl_gems)? If you are interested in reconstructing metabolic models directly from metagenomic samples then I can recommend the metaGEM pipeline, which uses CarveMe. In the paper we argue that metagenome-based metabolic models capture strain and context specific pangenomic variation that is missed by using reference genomes as a starting point of metabolic models 💎

Regarding your question about a reference database, we have deposited the models in a zenodo repository for now (https://zenodo.org/record/4407746#.YWFXFBbRYWM), but I do plan on making them more accessible (e.g. https://github.com/cdanielmachado/embl_gems) and also providing the underlying genomes + taxonomic annotation and other info/metadata generated by metaGEM.