Predict expression modulation targets¶
Cameo provides algorithms to search for genes or reactions that can be over- or down-regulated in order to achieve a given biological objective.
from cameo import models
from cameo.visualization.plotting.with_plotly import PlotlyPlotter
Load the E. coli core model.
model = models.bigg.e_coli_core
plotter = PlotlyPlotter()
Flux Scanning based on Enforced Objective Flux¶
from cameo.strain_design.deterministic.flux_variability_based import FSEOF
fseof = FSEOF(model)
fseof.run(target=model.reactions.EX_succ_e)
Reaction fluxes
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | |
|---|---|---|---|---|---|---|---|---|---|---|
| PFK | 7.756941 | 8.036499 | 8.242515 | 8.361048 | 8.479582 | 8.598115 | 8.747793 | 9.393102 | 9.830787 | 9.910940 |
| PGI | 5.485630 | 6.110399 | 6.513173 | 6.651876 | 6.790579 | 6.929282 | 7.161108 | 8.874882 | 9.965126 | 9.981645 |
| PGK | -16.431132 | -16.838737 | -17.173618 | -17.421989 | -17.670359 | -17.918730 | -18.198431 | -18.976723 | -19.547784 | -19.761991 |
| PGM | -15.238892 | -15.761645 | -16.212407 | -16.577535 | -16.942662 | -17.307790 | -17.704415 | -18.602292 | -19.293292 | -19.628048 |
| PPC | 3.758320 | 5.012330 | 5.878875 | 6.284516 | 6.690157 | 7.095799 | 7.532996 | 8.472367 | 9.615323 | 11.108854 |
| PPS | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.410338 | 1.527283 |
| ADK1 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.410338 | 1.527283 |
| ENO | 15.238892 | 15.761645 | 16.212407 | 16.577535 | 16.942662 | 17.307790 | 17.704415 | 18.602292 | 19.293292 | 19.628048 |
| SUCCt3 | 1.474575 | 2.949150 | 4.423725 | 5.898300 | 7.372875 | 8.847450 | 10.322025 | 11.796600 | 13.271175 | 14.745750 |
| THD2 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.000000 | 0.119674 | 2.143848 | 2.916807 | 1.535161 |
| TPI | 7.756941 | 8.036499 | 8.242515 | 8.361048 | 8.479582 | 8.598115 | 8.747793 | 9.393102 | 9.830787 | 9.910940 |
| FBA | 7.756941 | 8.036499 | 8.242515 | 8.361048 | 8.479582 | 8.598115 | 8.747793 | 9.393102 | 9.830787 | 9.910940 |
| FRD7 | 0.000000 | 0.000000 | 0.000000 | 1.453235 | 3.175302 | 4.897369 | 6.586703 | 7.755141 | 9.127842 | 10.852285 |
| GAPD | 16.431132 | 16.838737 | 17.173618 | 17.421989 | 17.670359 | 17.918730 | 18.198431 | 18.976723 | 19.547784 | 19.761991 |
| ICL | 0.000000 | 0.000000 | 0.386057 | 1.231342 | 2.076627 | 2.921912 | 3.735322 | 4.041459 | 4.143333 | 3.893465 |
| MALS | 0.000000 | 0.000000 | 0.386057 | 1.231342 | 2.076627 | 2.921912 | 3.735322 | 4.041459 | 4.143333 | 3.893465 |
| PDH | 8.809270 | 8.336008 | 8.179868 | 8.400957 | 8.622045 | 8.843134 | 9.064539 | 9.290984 | 9.107760 | 8.219084 |
| EX_succ_e | 1.474575 | 2.949150 | 4.423725 | 5.898300 | 7.372875 | 8.847450 | 10.322025 | 11.796600 | 13.271175 | 14.745750 |
Differential flux variability analysis¶
Compares flux ranges of a reference model to a set of models that have been parametrized to lie on a grid of evenly spaced points in the n-dimensional production envelope (n being the number of reaction bounds to be varied).
from cameo.flux_analysis.analysis import phenotypic_phase_plane
from cameo.strain_design.deterministic import DifferentialFVA
Succinate production¶
The production envelope looks like this.
production_envelope = phenotypic_phase_plane(model,
variables=[model.reactions.BIOMASS_Ecoli_core_w_GAM],
objective=model.metabolites.succ_e)
production_envelope.plot(plotter, height=400)
Set up a model that represents a reference state (in this case a model with a constrained growth rate).
model.reactions.EX_o2_e.lower_bound = 0
reference_model = model.copy()
biomass_rxn = reference_model.reactions.BIOMASS_Ecoli_core_w_GAM
biomass_rxn.lower_bound = 0.
target = reference_model.metabolites.succ_e
Set up the differential flux variability analysis strain design method.
diffFVA = DifferentialFVA(design_space_model=model,
reference_model=reference_model,
objective=target,
variables=[biomass_rxn],
normalize_ranges_by=biomass_rxn,
points=10)
Run differential flux variability analysis (only on the surface of the production envelope)
result = diffFVA.run(surface_only=True)
HBox()
result.solutions
| lower_bound | upper_bound | gaps | normalized_gaps | biomass | production | KO | flux_reversal | suddenly_essential | free_flux | reaction | excluded | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| reaction | ||||||||||||
| ACALD | 0.000000 | 0.000000 | -8.279455 | NaN | 0.000000 | 13.905778 | True | False | False | False | ACALD | False |
| ACALDt | 0.000000 | 0.000000 | 0.000000 | NaN | 0.000000 | 13.905778 | True | False | False | False | ACALDt | False |
| ACKr | -5.664889 | -5.664889 | -2.838696 | NaN | 0.000000 | 13.905778 | False | False | False | False | ACKr | False |
| ACONTa | 0.429333 | 0.429333 | 0.200970 | NaN | 0.000000 | 13.905778 | False | False | False | False | ACONTa | False |
| ACONTb | 0.429333 | 0.429333 | 0.200970 | NaN | 0.000000 | 13.905778 | False | False | False | False | ACONTb | False |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| TALA | -0.033659 | -0.033659 | -0.004207 | 0.000000 | 0.188145 | 2.123333 | False | False | False | False | TALA | False |
| THD2 | 3.225950 | 3.225950 | -0.403244 | 0.000000 | 0.188145 | 2.123333 | False | False | False | False | THD2 | False |
| TKT1 | -0.033659 | -0.033659 | -0.004207 | 0.000000 | 0.188145 | 2.123333 | False | False | False | False | TKT1 | False |
| TKT2 | -0.101579 | -0.101579 | -0.012697 | 0.000000 | 0.188145 | 2.123333 | False | False | False | False | TKT2 | False |
| TPI | 9.812852 | 9.812852 | 0.023393 | 5.905616 | 0.188145 | 2.123333 | False | False | False | False | TPI | False |
684 rows × 12 columns
result.plot(plotter, 5, variables=['FBP', 'G6PDH2r', 'PGL', 'PGK'])
result.display_on_map(2, map_name="iJO1366.Central metabolism")
Downloading Map from https://escher.github.io/1-0-0/6/maps/Escherichia%20coli/iJO1366.Central%20metabolism.json
Builder(reaction_data={'ACALD': -39.11622412349536, 'ACKr': 115.52882712367142, 'ACt2r': 115.52882712367142, '…
