Fix Duplicate Stearoyl-CoA Beta-Oxidation Reactions

[Devlin-Moyer]

Main improvements in this PR:

As proposed in #802:

• Removes MAR00776 for being a duplicate of MAR03115 + MAR03116 + MAR03117 + MAR03118
• Removes MAR00778 for being a duplicate of MAR03116 + MAR03117 + MAR03118 + MAR03121 + MAR03122 + MAR03123 + MAR03125 + MAR03128 + MAR03129 + MAR03130 + MAR03132 +MAR03135 + MAR03136 + MAR03137 + MAR03139 + MAR03142 + MAR03143 + MAR03144 + MAR03146

I noticed that MAR00776 had references associated with it but MAR03115, MAR03116, MAR03117, and MAR03118 didn't, but the two references associated with MAR00776 were about ACADS and ACADM, neither of which can actually catalyze the first round of beta-oxidation of stearoyl-CoA (at 18 carbons, it's too long), so I instead associated those four reactions with PMID:20490924, a review of mitochondrial beta-oxidation. MAR03121 onwards all had several references already associated with them, so I left those alone.

I hereby confirm that I have:

• Tested my code on my own computer for running the model
• Selected develop as a target branch
• Any removed reactions and metabolites have been moved to the corresponding deprecated identifier lists

[Devlin-Moyer]

it would be easier to figure out what went wrong with the metabolic tasks if the test indicated which tasks failed rather than just saying that the test as a whole failed (if this information is in the output somewhere and I'm just not seeing it, let me know)

[feiranl]

@JHL-452b Could you help check this?

[JHL-452b]

it would be easier to figure out what went wrong with the metabolic tasks if the test indicated which tasks failed rather than just saying that the test as a whole failed (if this information is in the output somewhere and I'm just not seeing it, let me know)

Yeah. In the current code, it would exit the matlab program directly if the task failed without any feedback. This suggestion should be incorporated into plans for subsequent improvements.

However, in this bug, I tested on my own computer and finded that the task of Growth on Ham's media (biomass production) failed. The cause of the failure is being analyzed, and will provide feedback as soon as possible when there are conclusions.

[JHL-452b]

Actually, the failed task seems to have nothing to do with the modification of this PR due to same problem occured when I re-run the task for ./develop/Human-GEM.

I tried changing the solver to gurobi, and all tasks succeeded as you can see below:

>> taskType = 'essential';
>> setRavenSolver('glpk')
>> status = testMetabolicTasks(taskType)
Boundary metabolites were added to 1660 reactions.
New (boundary) versions of 1660 metabolites were added to the model.

FAIL: [GR] Growth on Ham's media (biomass production)
status =
     0

>> setRavenSolver('gurobi')
>> status = testMetabolicTasks(taskType)
Boundary metabolites were added to 1660 reactions.
New (boundary) versions of 1660 metabolites were added to the model.

Suceeded with essential tasks.
status =
     1

I don't know why RAVEN cannot automatically change the solver to gurobi in the running code of those tasks now. If we don't resolve it, all subsequent PR will fail. If there is no problem with the recently updated content, can we directly add setRavenSolver('gurobi') to the code of tasks? @feiranl

[edkerk]

See #821 (comment) for consideration regarding gurobi/glpk.

[github-actions]

This PR has been automatically tested with GH Actions. Here is the output of the MACAW test:

Starting dead-end test...
 - Found 1514 dead-end metabolites.
 - Found 1319 reactions incapable of sustaining steady-state fluxes in either direction due to these dead-ends.
 - Found 1977 reversible reactions that can only carry steady-state fluxes in a single direction due to dead-ends.
Starting duplicate test...
 - Skipping redox duplicates because no redox_pairs and/or proton_ids were provided.
 - Found 447 reactions that were some type of duplicate:
   - 0 were completely identical to at least one other reaction.
   - 13 involve the same metabolites but go in the opposite direction or have the opposite reversibility as at least one other reaction.
   - 447 involve the same metabolites but with different coefficients as at least one other reaction.

This and a more detailed output from MACAW are also committed to data/macawResults/.

Note: In the case of multiple test runs, this post will be edited.

[feiranl]

@Devlin-Moyer @edkerk Is this PR ready to be merged?

[edkerk]

Yes, but I'd like to merge-in develop when #883 is approved.

[github-actions]

This PR has been automatically tested with GH Actions. Here is the output of the gene essentiality test:

     cellLine     TP     TN     FP    FN     accuracy    sensitivity    specificity       F1         MCC   
    __________    __    ____    __    ___    ________    ___________    ___________    ________    ________

    {'DLD1'  }    36    2181    61    279    0.86703       0.11429        0.97279       0.17476     0.14981
    {'GBM'   }    34    2161    63    298    0.85876       0.10241        0.97167       0.15851     0.13034
    {'HCT116'}    45    2202    56    310    0.85993       0.12676         0.9752       0.19737     0.18123
    {'HELA'  }    30    2260    71    253    0.87605       0.10601        0.96954       0.15625      0.1218
    {'RPE1'  }    15    2201    82    258    0.86698      0.054945        0.96408      0.081081    0.030757
    {'all'   }     7    2404    94    109    0.92234      0.060345        0.96237      0.064516     0.02427

Note: In the case of multiple test runs, this post will be edited.