fluxbound

A minimalistic python package for flux balance analysis

Installation

pip install fluxbound

Solvers

SCIP solver is installed by default. You can optionally install GUROBI or CPLEX.

If multiple solvers are installed, the selection order is: GUROBI > CPLEX > SCIP.

But you can change your prefered solver:

from fluxbound import set_default_solver

set_default_solver('scip')

Basic usage

Loading

Loading (and inspecting) a model:

from fluxbound import load_model

model = load_model('e_coli_core.xml')
model

Model: e_coli_core
  - Genes: 137
  - Metabolites: 72
      c: 52
      e: 20
  - Reactions: 95
      internal: 50
      transport: 25
      exchange: 20
  - Obj: R_BIOMASS_Ecoli_core_w_GAM

FBA

Running a basic FBA simulation:

from fluxbound import FBA

sol = FBA(model)
print(sol)

Objective: 0.8739
Status: Optimal

You can switch to pFBA:

sol = FBA(model, parsimonious=True)

You can add additional constraints:

constraints = {
    'R_EX_glc__D_e': (-5, 0),
    'R_EX_o2_e': 0,
}

sol = FBA(model, constraints=constraints)

And easily change the objective:

sol = FBA(model, objective='R_ATPM', minimize=True)

Solution

The solution object has utilities such a filtering with regular expressions:

sol.show_values(pattern='R_EX', sort=True, abstol=0.1)

R_EX_o2_e    -21.8
R_EX_glc__D_e -10
R_EX_nh4_e   -4.765
R_EX_pi_e    -3.215
R_EX_h_e      17.53
R_EX_co2_e    22.81
R_EX_h2o_e    29.18

And converting to pandas DataFrames

df = sol.to_dataframe()
df.query('value > 30')

              value
R_ATPS4r  45.514010
R_CYTBD   43.598985
R_NADH16  38.534610

FVA

Running Flux Variability Analysis:

from fluxbound import FVA

ranges = FVA(model)

FVA at 90% maximum growth in anaerobic conditions:

ranges = FVA(
    model, 
    obj_frac = 0.9,
    constraints = {'R_EX_o2_e': 0},
)

Deletion

You can simulate the deletion of multiple genes:

from fluxbound import gene_deletion

sol = gene_deletion(model, ['G_b0008', 'G_b1241', 'G_b4301'])

Or reactions:

from fluxbound import reaction_deletion

sol = reaction_deletion(model, ['R_PGI', 'R_PFK'])

Essentiality

Calculate all essential genes or reactions:

from fluxbound import essential_genes

genes = essential_genes(model)

Remember: you can always use the constraints argument to change conditions:

from fluxbound import essential_reactions

reactions = essential_reactions(model, constraints={'R_EX_o2_e': 0})

Advanced

Environments

The Environment class is an utility for easily creating and changing environmental conditions.

You can create a temporary environment to use in a simulation:

from fluxbound import Environment

env = Environment.complete(model, inplace=False)
sol = FBA(model, constraints=env)

Or permanently change a model:

Environment.complete(model, inplace=True)
sol = FBA(model)

Environments are dictionary objects that you can easily manipulate:

env = Environment.from_model(model)
env['R_EX_o2_e'] = (-20, 0)
env.simplify()

{'R_EX_co2_e': (-1000.0, 1000.0),
 'R_EX_glc__D_e': (-10.0, 1000.0),
 'R_EX_h_e': (-1000.0, 1000.0),
 'R_EX_h2o_e': (-1000.0, 1000.0),
 'R_EX_nh4_e': (-1000.0, 1000.0),
 'R_EX_o2_e': (-20, 0),
 'R_EX_pi_e': (-1000.0, 1000.0)}

Model manipulation

If instead of just simulating some environmental or genetic perturbations you want to actually modify a model, you can play around with the model object.

Make changes:

model.reactions.R_PGI.lb = 0
model.set_flux_bounds('R_ATPM', lb=1, ub=1)

Add elements:

from fluxbound import Metabolite, Reaction

x = Metabolite('M_x_e', name='compound x', compartment='e')
model.add_metabolite(x)
r = Reaction('R_EX_x_e', stoichiometry={'M_x_e': -1}, lb=0, ub=1000)
model.add_reaction(r)
print(model.reactions.R_EX_y_e)

R_EX_y_e: M_x_e -->  [0, 1000]

Remove elements:

model.remove_reactions(
    ['R_EX_fru_e', 'R_FRUpts2'],
    clean_orphans=True
) # removes orphaned metabolites and genes 

Finally you can save your new model as SBML:

from fluxbound import save_model

save_model(model, 'new_model.xml')