Tutorial on metabolic modelling

Using conda

The following has to be done once

# load the module
module load conda

conda init

To activate an environment:

conda activate my_env

To deactivate an environment:

conda deactivate

Paths to the conda environments for this tutorial

/shared/projects/tp_2534_ai_microbiomes_181502/conda/envs/ebame_metabo_gapseq
/shared/projects/tp_2534_ai_microbiomes_181502/conda/envs/ebame_metabo_reasoning
/shared/projects/tp_2534_ai_microbiomes_181502/conda/envs/numerical_modelling

We will mostly be using /shared/projects/tp_2534_ai_microbiomes_181502/conda/envs/ebame_metabo_reasoning

Tutorial

See https://gitlab.inria.fr/cfrioux/ebame for the complete tutorial. The tutorial has a section dedicated to metabolic network reconstruction which we may not consider during this training if time runs short.

To start, get the data by cloning the git repository:

git clone https://gitlab.inria.fr/cfrioux/ebame.git

And activate the following environment:

conda activate /shared/projects/tp_2534_ai_microbiomes_181502/conda/envs/ebame_metabo_reasoning

We will do the following:

• Manipulation of metabolic models with Fluxer
• A closer look at a metabolic network: the SBML file
• Metabolic network reconstruction with gapseq from the protein sequences, or with Kegg2bipartite from protein annotations
• Metabolic network modelling
  • Using toy data
  • Using real networks
• Screening the metabolism of microbial communities
  • Using toy community data
  • Using realistic communities
• BONUS - Inferring seed metabolites (nutrients) from the metabolic network

Wanna play with Answer Set Programming?

reaction("r1").
reactant("A", "r1").
product("B", "r1").

reaction("r2").
reactant("B", "r2").
reactant("C", "r2").
product("D", "r2").

seed("A").
seed("C").

scope(M) :- seed(M).
scope(M) :- product(M,R); reaction(R); scope(N) : reactant(N,R).

#show scope/1.

Try it on the clingo online solver