From 98ca9f4e5689e2f0afcab15dbe4abdb4d8504ae0 Mon Sep 17 00:00:00 2001 From: Eduard Kerkhoven Date: Fri, 20 Nov 2026 21:12:54 +0100 Subject: [PATCH 1/2] refactor: use RAVEN's parseRAVENargs and remove parseGECKOargs parseRAVENargs accepts everything parseGECKOargs did and also leading positional arguments followed by name-value pairs. Only the error identifiers change, from parseGECKOargs:* to parseRAVENargs:*. --- src/change_model/addNewRxnsToEC.m | 2 +- src/change_model/applyComplexData.m | 2 +- src/change_model/applyKcatConstraints.m | 2 +- src/change_model/findMetSmiles.m | 2 +- src/change_model/getComplexData.m | 2 +- src/change_model/makeEcModel.m | 2 +- src/change_model/setProtPoolSize.m | 2 +- src/deprecated/applyCustomKcats.m | 2 +- src/deprecated/mergeDLKcatAndFuzzyKcats.m | 2 +- src/deprecated/readDLKcatOutput.m | 2 +- src/deprecated/updateProtPool.m | 2 +- src/download_databases/loadBRENDAdata.m | 2 +- src/download_databases/loadDatabases.m | 2 +- src/enzyme_usage/getEnzymeBottlenecks.m | 2 +- src/enzyme_usage/getEnzymeUsage.m | 2 +- src/enzyme_usage/reportEnzymeUsage.m | 2 +- src/flux_analysis/ecFSEOF.m | 2 +- src/flux_analysis/ecFVA.m | 2 +- src/flux_analysis/getPfbaEnzymes.m | 2 +- src/gather_kcats/assignKcatValues.m | 2 +- src/gather_kcats/collectKcats.m | 2 +- src/gather_kcats/convertCustomKcats.m | 2 +- .../extractEnzymeSubstratePairs.m | 2 +- src/gather_kcats/fuzzyKcatMatching.m | 4 +- src/gather_kcats/getStandardKcat.m | 2 +- src/gather_kcats/kcatsFromBrenda.m | 2 +- src/gather_kcats/kcatsFromDLKcat.m | 8 +- src/gather_kcats/kcatsFromECTree.m | 2 +- src/gather_kcats/kcatsFromHomologs.m | 2 +- src/gather_kcats/kcatsFromOKP.m | 6 +- src/gather_kcats/loadBrendaMeasurements.m | 2 +- src/gather_kcats/loadKcatSource.m | 2 +- src/gather_kcats/mergeKcats.m | 2 +- src/gather_kcats/reviewKcatAssignment.m | 2 +- src/gather_kcats/saveKcatSource.m | 2 +- src/gather_kcats/updateKcatSources.m | 2 +- src/get_enzyme_data/applyECcodes.m | 2 +- src/get_enzyme_data/getECfromDatabase.m | 2 +- src/get_enzyme_data/getECfromGEM.m | 2 +- src/io/loadConventionalGEM.m | 2 +- src/io/loadEcModel.m | 2 +- src/io/loadFluxData.m | 2 +- src/io/loadProtData.m | 2 +- src/io/saveEcModel.m | 2 +- src/kcat_tuning/evotune/bestParsimonious.m | 2 +- src/kcat_tuning/evotune/cmaesKcatTuning.m | 2 +- src/kcat_tuning/evotune/evotuneScore.m | 2 +- src/kcat_tuning/evotune/isozymeTieMap.m | 2 +- src/kcat_tuning/evotune/loadEvotuneData.m | 2 +- src/kcat_tuning/evotune/nChanged.m | 2 +- src/kcat_tuning/evotune/parsimonyFrontier.m | 2 +- src/kcat_tuning/evotune/screenKcatLeverage.m | 2 +- src/kcat_tuning/evotune/selectTunableMask.m | 2 +- .../evotune/tunePriorPenaltyWeight.m | 2 +- src/kcat_tuning/sensitivityTuning.m | 2 +- src/kcat_tuning/sigmaFitter.m | 2 +- src/limit_proteins/calculateFfactor.m | 2 +- src/limit_proteins/constrainEnzConcs.m | 2 +- src/limit_proteins/constrainFluxData.m | 2 +- src/limit_proteins/fillEnzConcs.m | 2 +- src/limit_proteins/flexibilizeEnzConcs.m | 2 +- src/limit_proteins/getComplexFamilies.m | 2 +- src/limit_proteins/getConcControlCoeffs.m | 2 +- src/limit_proteins/levelComplexConcs.m | 2 +- src/limit_proteins/relaxEnzConcs.m | 2 +- src/limit_proteins/relaxProteomicsGreedy.m | 2 +- src/limit_proteins/zeroUnmeasuredIsozymes.m | 2 +- src/utilities/createGECKOproject.m | 2 +- src/utilities/parseGECKOargs.m | 98 ------------------- src/utilities/upgradeGECKOproject.m | 2 +- test/unit_tests/geckoCoreFunctionTests.m | 9 +- 71 files changed, 79 insertions(+), 178 deletions(-) delete mode 100644 src/utilities/parseGECKOargs.m diff --git a/src/change_model/addNewRxnsToEC.m b/src/change_model/addNewRxnsToEC.m index f9ef29aa..58ee4710 100644 --- a/src/change_model/addNewRxnsToEC.m +++ b/src/change_model/addNewRxnsToEC.m @@ -67,7 +67,7 @@ % -------- % applyKcatConstraints, setKcatForReactions, assignKcats -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'newEnzymes', []; ... 'modelAdapter', []}); newEnzymes = p.newEnzymes; diff --git a/src/change_model/applyComplexData.m b/src/change_model/applyComplexData.m index 7f024208..23e0465d 100644 --- a/src/change_model/applyComplexData.m +++ b/src/change_model/applyComplexData.m @@ -60,7 +60,7 @@ % -------- % getComplexData, makeEcModel -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'complexInfo', []; ... 'modelAdapter', []; ... 'verbose', true}); diff --git a/src/change_model/applyKcatConstraints.m b/src/change_model/applyKcatConstraints.m index 6adc7c26..f8d12ff6 100644 --- a/src/change_model/applyKcatConstraints.m +++ b/src/change_model/applyKcatConstraints.m @@ -41,7 +41,7 @@ % -------- % makeEcModel, fillKcatsFromIsozymes, setKcatForReactions -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'updateRxns', []}); updateRxns = p.updateRxns; diff --git a/src/change_model/findMetSmiles.m b/src/change_model/findMetSmiles.m index 364d4a45..4a389202 100644 --- a/src/change_model/findMetSmiles.m +++ b/src/change_model/findMetSmiles.m @@ -27,7 +27,7 @@ % noSMILES : cell % metabolite names for which no SMILES could be found. % -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... 'verbose', true}); modelAdapter = p.modelAdapter; diff --git a/src/change_model/getComplexData.m b/src/change_model/getComplexData.m index cf0541a4..bfc48af0 100644 --- a/src/change_model/getComplexData.m +++ b/src/change_model/getComplexData.m @@ -46,7 +46,7 @@ % -------- % applyComplexData, makeEcModel -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'taxonomicID', []; ... 'modelAdapter', []}); taxonomicID = p.taxonomicID; diff --git a/src/change_model/makeEcModel.m b/src/change_model/makeEcModel.m index 5ae64050..e4a1d92d 100644 --- a/src/change_model/makeEcModel.m +++ b/src/change_model/makeEcModel.m @@ -110,7 +110,7 @@ % -------- % applyKcatConstraints, applyComplexData -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'geckoLight', false; ... 'modelAdapter', []}); geckoLight = p.geckoLight; diff --git a/src/change_model/setProtPoolSize.m b/src/change_model/setProtPoolSize.m index 440563f8..4825b7bd 100644 --- a/src/change_model/setProtPoolSize.m +++ b/src/change_model/setProtPoolSize.m @@ -43,7 +43,7 @@ % -------- % makeEcModel -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'Ptot', []; ... 'f', []; ... 'sigma', []; ... diff --git a/src/deprecated/applyCustomKcats.m b/src/deprecated/applyCustomKcats.m index 592da03d..9dd4906a 100644 --- a/src/deprecated/applyCustomKcats.m +++ b/src/deprecated/applyCustomKcats.m @@ -88,7 +88,7 @@ warning('GECKO:deprecatedFunction', ['applyCustomKcats is deprecated; use convertCustomKcats, ' ... 'which writes data/kcats/custom.tsv for collectKcats and assignKcats. applyCustomKcats ' ... 'will be removed in a future release.']) -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'customKcats', []; ... 'modelAdapter', []}); customKcats = p.customKcats; diff --git a/src/deprecated/mergeDLKcatAndFuzzyKcats.m b/src/deprecated/mergeDLKcatAndFuzzyKcats.m index 6610e93a..2ac4f634 100644 --- a/src/deprecated/mergeDLKcatAndFuzzyKcats.m +++ b/src/deprecated/mergeDLKcatAndFuzzyKcats.m @@ -74,7 +74,7 @@ % -------- % mergeKcats, readDLKcatOutput, fuzzyKcatMatching, assignKcatValues -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'topOriginLimit', 4; ... 'bottomOriginLimit', 6; ... 'wildcardLimit', 3}); diff --git a/src/deprecated/readDLKcatOutput.m b/src/deprecated/readDLKcatOutput.m index 04dce061..fa20b8a7 100644 --- a/src/deprecated/readDLKcatOutput.m +++ b/src/deprecated/readDLKcatOutput.m @@ -21,7 +21,7 @@ warning('GECKO:deprecatedFunction', ['readDLKcatOutput is deprecated; use ' ... 'kcatsFromDLKcat(model, ''via'', ''file''). It will be removed in a future release.']) -p = parseGECKOargs(varargin, {'outFile', []; 'modelAdapter', []; 'strict', []}); +p = parseRAVENargs(varargin, {'outFile', []; 'modelAdapter', []; 'strict', []}); if ~isempty(p.strict) warning('GECKO:deprecatedOption', ['The strict option of readDLKcatOutput is ignored: rows ' ... 'with a substrate that is not in model.metNames are left out with a warning.']) diff --git a/src/deprecated/updateProtPool.m b/src/deprecated/updateProtPool.m index fa0037d9..520bc4e5 100644 --- a/src/deprecated/updateProtPool.m +++ b/src/deprecated/updateProtPool.m @@ -49,7 +49,7 @@ warning('GECKO:deprecatedFunction', ['updateProtPool is deprecated; use setProtPoolSize ' ... 'instead. The function will be removed in a future release.']); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'Ptot', []; ... 'modelAdapter', []}); Ptot = p.Ptot; diff --git a/src/download_databases/loadBRENDAdata.m b/src/download_databases/loadBRENDAdata.m index d5753fab..672a0a44 100644 --- a/src/download_databases/loadBRENDAdata.m +++ b/src/download_databases/loadBRENDAdata.m @@ -51,7 +51,7 @@ % -------- % loadDatabases, getECfromDatabase -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'modelAdapter', [], @(x) assert(isempty(x) || isa(x,'ModelAdapter'), ... 'loadBRENDAdata:invalidModelAdapter', ... 'modelAdapter must be a ModelAdapter.'); ... diff --git a/src/download_databases/loadDatabases.m b/src/download_databases/loadDatabases.m index 9abdf6ea..843246a4 100644 --- a/src/download_databases/loadDatabases.m +++ b/src/download_databases/loadDatabases.m @@ -35,7 +35,7 @@ % -------- % getECfromDatabase, loadBRENDAdata, downloadKEGG, downloadUniProt -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'selectDatabase', 'both'; ... 'modelAdapter', []}); selectDatabase = p.selectDatabase; diff --git a/src/enzyme_usage/getEnzymeBottlenecks.m b/src/enzyme_usage/getEnzymeBottlenecks.m index 8b68e11b..02bf6f0f 100644 --- a/src/enzyme_usage/getEnzymeBottlenecks.m +++ b/src/enzyme_usage/getEnzymeBottlenecks.m @@ -56,7 +56,7 @@ % bottlenecks = getEnzymeBottlenecks(ecModel); % bottlenecks = getEnzymeBottlenecks(ecModel, 'top', 5, 'foldChange', 3); -p = parseGECKOargs(varargin, {'top', []; 'foldChange', []}); +p = parseRAVENargs(varargin, {'top', []; 'foldChange', []}); top = p.top; foldChange = p.foldChange; if isempty(top) diff --git a/src/enzyme_usage/getEnzymeUsage.m b/src/enzyme_usage/getEnzymeUsage.m index c48017c9..b3507411 100644 --- a/src/enzyme_usage/getEnzymeUsage.m +++ b/src/enzyme_usage/getEnzymeUsage.m @@ -74,7 +74,7 @@ % -------- % reportEnzymeUsage -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'zero', true}); zero = p.zero; if ecModel.ec.geckoLight diff --git a/src/enzyme_usage/reportEnzymeUsage.m b/src/enzyme_usage/reportEnzymeUsage.m index 6a862b45..db101186 100644 --- a/src/enzyme_usage/reportEnzymeUsage.m +++ b/src/enzyme_usage/reportEnzymeUsage.m @@ -50,7 +50,7 @@ % -------- % getEnzymeUsage -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'highCapUsage', []; ... 'topAbsUsage', []}); highCapUsage = p.highCapUsage; diff --git a/src/flux_analysis/ecFSEOF.m b/src/flux_analysis/ecFSEOF.m index eb80cb49..f51d2b15 100644 --- a/src/flux_analysis/ecFSEOF.m +++ b/src/flux_analysis/ecFSEOF.m @@ -71,7 +71,7 @@ % fseof = ecFSEOF(model,prodTargetRxn,csRxn,nSteps,outputFile,filePath,modelAdapter,corrThreshold); % fseof = ecFSEOF(model,prodTargetRxn,csRxn,'nSteps',32); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'nSteps', 16; ... 'outputFile', false; ... 'filePath', []; ... diff --git a/src/flux_analysis/ecFVA.m b/src/flux_analysis/ecFVA.m index ae624b24..2045a0ca 100644 --- a/src/flux_analysis/ecFVA.m +++ b/src/flux_analysis/ecFVA.m @@ -38,7 +38,7 @@ % -------- % mapRxnsToConv, plotEcFVA -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'runParallel', true}); runParallel = p.runParallel; diff --git a/src/flux_analysis/getPfbaEnzymes.m b/src/flux_analysis/getPfbaEnzymes.m index 68dc29d7..f5ccfd99 100644 --- a/src/flux_analysis/getPfbaEnzymes.m +++ b/src/flux_analysis/getPfbaEnzymes.m @@ -51,7 +51,7 @@ % -------- % getEnzymeBottlenecks -p = parseGECKOargs(varargin, {'rxnId', []; 'fractionOfOptimum', []}); +p = parseRAVENargs(varargin, {'rxnId', []; 'fractionOfOptimum', []}); rxnId = p.rxnId; fractionOfOptimum = p.fractionOfOptimum; if isempty(fractionOfOptimum) diff --git a/src/gather_kcats/assignKcatValues.m b/src/gather_kcats/assignKcatValues.m index f3c45fbf..4d977e96 100644 --- a/src/gather_kcats/assignKcatValues.m +++ b/src/gather_kcats/assignKcatValues.m @@ -65,7 +65,7 @@ % -------- % kcatsFromDLKcat, mergeKcats, assignKcats, applyKcatConstraints -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'criteria', []; ... 'overwrite', []}); criteria = p.criteria; diff --git a/src/gather_kcats/collectKcats.m b/src/gather_kcats/collectKcats.m index ce46269a..595a1373 100644 --- a/src/gather_kcats/collectKcats.m +++ b/src/gather_kcats/collectKcats.m @@ -121,7 +121,7 @@ % assignKcats, updateKcatSources, loadKcatSource, mergeKcats, assignKcatValues checkOptions(varargin, {'modelAdapter','writeReport'}, 'collectKcats'); -p = parseGECKOargs(varargin, {'modelAdapter', []; 'writeReport', true}); +p = parseRAVENargs(varargin, {'modelAdapter', []; 'writeReport', true}); modelAdapter = p.modelAdapter; if isempty(modelAdapter) modelAdapter = ModelAdapterManager.getDefault(); diff --git a/src/gather_kcats/convertCustomKcats.m b/src/gather_kcats/convertCustomKcats.m index a79d593b..3ad81032 100644 --- a/src/gather_kcats/convertCustomKcats.m +++ b/src/gather_kcats/convertCustomKcats.m @@ -43,7 +43,7 @@ % collectKcats, applyCustomKcats, getEcRxnKeys, upgradeGECKOproject checkOptions(varargin, {'modelAdapter','customKcats','outFile'}, 'convertCustomKcats'); -p = parseGECKOargs(varargin, {'modelAdapter', []; 'customKcats', []; 'outFile', []}); +p = parseRAVENargs(varargin, {'modelAdapter', []; 'customKcats', []; 'outFile', []}); modelAdapter = p.modelAdapter; if isempty(modelAdapter) modelAdapter = ModelAdapterManager.getDefault(); diff --git a/src/gather_kcats/extractEnzymeSubstratePairs.m b/src/gather_kcats/extractEnzymeSubstratePairs.m index 094f0782..6e329509 100644 --- a/src/gather_kcats/extractEnzymeSubstratePairs.m +++ b/src/gather_kcats/extractEnzymeSubstratePairs.m @@ -39,7 +39,7 @@ % kcatsFromDLKcat, kcatsFromHomologs [geckoPath, ~] = findGECKOroot(); -p = parseGECKOargs(varargin, {'ecRxns', []; 'modelAdapter', []}); +p = parseRAVENargs(varargin, {'ecRxns', []; 'modelAdapter', []}); ecRxnsSel = p.ecRxns; if isempty(ecRxnsSel) ecRxnsSel = true(numel(model.ec.rxns),1); diff --git a/src/gather_kcats/fuzzyKcatMatching.m b/src/gather_kcats/fuzzyKcatMatching.m index ba7eefa8..07de8c0a 100644 --- a/src/gather_kcats/fuzzyKcatMatching.m +++ b/src/gather_kcats/fuzzyKcatMatching.m @@ -88,10 +88,10 @@ % mergeKcats, assignKcats, assignKcatValues % Validators reject a name that landed in a positional slot, which -% happens when positional and name-value arguments are mixed. Without +% happens when a positional value follows name-value pairs. Without % them `while forceWClvl > 0` on a char array is all-true and silently % wildcards every EC number. -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'ecRxns', [], @(x) assert(isempty(x) || isnumeric(x) || ... islogical(x), ... 'fuzzyKcatMatching:invalidEcRxns', ... diff --git a/src/gather_kcats/getStandardKcat.m b/src/gather_kcats/getStandardKcat.m index badd9e0d..9e68a138 100644 --- a/src/gather_kcats/getStandardKcat.m +++ b/src/gather_kcats/getStandardKcat.m @@ -84,7 +84,7 @@ % -------- % removeStandardKcat, applyKcatConstraints -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... 'threshold', []; ... 'fillZeroKcat', []}); diff --git a/src/gather_kcats/kcatsFromBrenda.m b/src/gather_kcats/kcatsFromBrenda.m index de65b67e..662a111c 100644 --- a/src/gather_kcats/kcatsFromBrenda.m +++ b/src/gather_kcats/kcatsFromBrenda.m @@ -48,7 +48,7 @@ % fuzzyKcatMatching, kcatsFromDLKcat, kcatsFromOKP, saveKcatSource, loadKcatSource checkOptions(varargin, {'aggregation','forceWClvl','ecRxns','save','saveAs','modelAdapter'}, 'kcatsFromBrenda'); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'aggregation', []; ... 'forceWClvl', 0; ... 'ecRxns', []; ... diff --git a/src/gather_kcats/kcatsFromDLKcat.m b/src/gather_kcats/kcatsFromDLKcat.m index af855b69..bfbbbaf7 100644 --- a/src/gather_kcats/kcatsFromDLKcat.m +++ b/src/gather_kcats/kcatsFromDLKcat.m @@ -66,7 +66,7 @@ % kcatsFromOKP, extractEnzymeSubstratePairs, saveKcatSource checkOptions(varargin, {'via','run','wait','ecRxns','save','saveAs','modelAdapter'}, 'kcatsFromDLKcat'); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'via', []; ... 'run', false; ... 'wait', false; ... @@ -185,7 +185,7 @@ function checkOptions(args, names, functionName) function writtenTable = writeDLKcatInput(model, varargin) % Writes the DLKcat input (data/DLKcat.tsv by default): one row per % enzyme-substrate pair, with the kcat column NA. -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'ecRxns', []; ... 'modelAdapter', []; ... 'onlyWithSmiles', []; ... @@ -275,7 +275,7 @@ function writeRows(filename, out) function runDLKcat(varargin) % Runs DLKcat in its Docker image on data/DLKcat.tsv, which it overwrites % with the predictions. -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... 'filePath', []}); modelAdapter = p.modelAdapter; @@ -334,7 +334,7 @@ function runDLKcat(varargin) % (readDLKcatOutput:unknownSubstrate). A file without kcats raises % readDLKcatOutput:noKcats, a reaction that is not in model.ec.rxns % readDLKcatOutput:unknownReaction. -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'outFile', []; ... 'modelAdapter', []}); outFile = p.outFile; diff --git a/src/gather_kcats/kcatsFromECTree.m b/src/gather_kcats/kcatsFromECTree.m index d1906d83..5db6948c 100644 --- a/src/gather_kcats/kcatsFromECTree.m +++ b/src/gather_kcats/kcatsFromECTree.m @@ -53,7 +53,7 @@ % ecTreeEstimate, loadBrendaMeasurements, kcatsFromBrenda, saveKcatSource, assignKcats checkOptions(varargin, {'k','ecRxns','measurements','save','saveAs','modelAdapter'}, 'kcatsFromECTree'); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'k', 5; ... 'ecRxns', []; ... 'measurements', []; ... diff --git a/src/gather_kcats/kcatsFromHomologs.m b/src/gather_kcats/kcatsFromHomologs.m index 62bff5c8..5ab69dce 100644 --- a/src/gather_kcats/kcatsFromHomologs.m +++ b/src/gather_kcats/kcatsFromHomologs.m @@ -57,7 +57,7 @@ % kcatsFromECTree, extractEnzymeSubstratePairs, saveKcatSource, collectKcats checkOptions(varargin, {'k','minIdentity','coverage','threads','save','saveAs','modelAdapter'}, 'kcatsFromHomologs'); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'k', 10; ... 'minIdentity', 0.4; ... 'coverage', 0.8; ... diff --git a/src/gather_kcats/kcatsFromOKP.m b/src/gather_kcats/kcatsFromOKP.m index bca5248e..0949b188 100644 --- a/src/gather_kcats/kcatsFromOKP.m +++ b/src/gather_kcats/kcatsFromOKP.m @@ -78,7 +78,7 @@ % kcatsFromBrenda, kcatsFromDLKcat, resolveOkpApiKey, saveKcatSource checkOptions(varargin, {'method','wait','resubmit','useStored','ecRxns','save','saveAs','modelAdapter'}, 'kcatsFromOKP'); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'method', []; ... 'wait', false; ... 'resubmit', false; ... @@ -393,7 +393,7 @@ function writeOpenKineticsPredictorInput(model, varargin) [geckoPath, ~] = findGECKOroot(); %% Parse inputs -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'ecRxns', []; ... 'modelAdapter', []; ... 'onlyWithSmiles', []; ... @@ -544,7 +544,7 @@ function writeOpenKineticsPredictorInput(model, varargin) % 100 s^-1') is read as its geometric midpoint, or half the upper bound for % the class from 0. kcatSource keeps the OKP source of each value, with % the prefix 'OKP-'. -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'outFile', []; ... 'modelAdapter', []}); outFile = p.outFile; diff --git a/src/gather_kcats/loadBrendaMeasurements.m b/src/gather_kcats/loadBrendaMeasurements.m index 3023f4c7..dd68b00c 100644 --- a/src/gather_kcats/loadBrendaMeasurements.m +++ b/src/gather_kcats/loadBrendaMeasurements.m @@ -34,7 +34,7 @@ % kcatsFromECTree, ecTreeEstimate, loadBRENDAdata checkOptions(varargin, {'modelAdapter','folder'}, 'loadBrendaMeasurements'); -p = parseGECKOargs(varargin, {'modelAdapter', []; 'folder', []}); +p = parseRAVENargs(varargin, {'modelAdapter', []; 'folder', []}); folder = p.folder; if isempty(folder) modelAdapter = p.modelAdapter; diff --git a/src/gather_kcats/loadKcatSource.m b/src/gather_kcats/loadKcatSource.m index 51e2f5ce..e19be8e2 100644 --- a/src/gather_kcats/loadKcatSource.m +++ b/src/gather_kcats/loadKcatSource.m @@ -71,7 +71,7 @@ % saveKcatSource, kcatsFromBrenda, getEcRxnKeys checkOptions(varargin, {'modelAdapter'}, 'loadKcatSource'); -p = parseGECKOargs(varargin, {'modelAdapter', []}); +p = parseRAVENargs(varargin, {'modelAdapter', []}); if isfile(name) filename = name; diff --git a/src/gather_kcats/mergeKcats.m b/src/gather_kcats/mergeKcats.m index a6321d24..775f09c4 100644 --- a/src/gather_kcats/mergeKcats.m +++ b/src/gather_kcats/mergeKcats.m @@ -95,7 +95,7 @@ % -------- % assignKcats, kcatsFromDLKcat, fuzzyKcatMatching, assignKcatValues -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'topOriginLimit', 4; ... 'bottomOriginLimit', 6; ... 'wildcardLimit', 3; ... diff --git a/src/gather_kcats/reviewKcatAssignment.m b/src/gather_kcats/reviewKcatAssignment.m index 47360453..4e13a38f 100644 --- a/src/gather_kcats/reviewKcatAssignment.m +++ b/src/gather_kcats/reviewKcatAssignment.m @@ -75,7 +75,7 @@ % See also % -------- % ecStatsFromBrenda, isozymeTieMap, screenKcatLeverage -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'ecStats', containers.Map('KeyType','char','ValueType','any'); ... 'enzymes', {}; ... 'names', {}; ... diff --git a/src/gather_kcats/saveKcatSource.m b/src/gather_kcats/saveKcatSource.m index 59e65a67..0aed6bcb 100644 --- a/src/gather_kcats/saveKcatSource.m +++ b/src/gather_kcats/saveKcatSource.m @@ -67,7 +67,7 @@ names = {'modelAdapter','folder','generator','parameters','data','kind'}; checkOptions(varargin, names, 'saveKcatSource'); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... 'folder', []; ... 'generator', 'saveKcatSource'; ... diff --git a/src/gather_kcats/updateKcatSources.m b/src/gather_kcats/updateKcatSources.m index 8ca1d4b3..101803f2 100644 --- a/src/gather_kcats/updateKcatSources.m +++ b/src/gather_kcats/updateKcatSources.m @@ -60,7 +60,7 @@ % collectKcats, assignKcats, kcatsFromBrenda, kcatsFromDLKcat, kcatsFromOKP checkOptions(varargin, {'dryRun','modelAdapter'}, 'updateKcatSources'); -p = parseGECKOargs(varargin, {'dryRun', false; 'modelAdapter', []}); +p = parseRAVENargs(varargin, {'dryRun', false; 'modelAdapter', []}); modelAdapter = p.modelAdapter; if isempty(modelAdapter) modelAdapter = ModelAdapterManager.getDefault(); diff --git a/src/get_enzyme_data/applyECcodes.m b/src/get_enzyme_data/applyECcodes.m index 59a81ec7..aa516319 100644 --- a/src/get_enzyme_data/applyECcodes.m +++ b/src/get_enzyme_data/applyECcodes.m @@ -29,7 +29,7 @@ % ecModel = applyECcodes(ecModel, true); % ecModel = applyECcodes(ecModel, 'overwrite', true); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'overwrite', false}); overwrite = p.overwrite; diff --git a/src/get_enzyme_data/getECfromDatabase.m b/src/get_enzyme_data/getECfromDatabase.m index 2c6a1981..552c8947 100644 --- a/src/get_enzyme_data/getECfromDatabase.m +++ b/src/get_enzyme_data/getECfromDatabase.m @@ -47,7 +47,7 @@ % -------- % getECfromGEM, findECInDB, loadDatabases -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'ecRxns', []; ... 'action', 'display'; ... 'modelAdapter', []}); diff --git a/src/get_enzyme_data/getECfromGEM.m b/src/get_enzyme_data/getECfromGEM.m index 6db41fe1..7ad2c49b 100644 --- a/src/get_enzyme_data/getECfromGEM.m +++ b/src/get_enzyme_data/getECfromGEM.m @@ -44,7 +44,7 @@ % -------- % getECfromDatabase, applyECcodes -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'ecRxns', []}); ecRxns = p.ecRxns; diff --git a/src/io/loadConventionalGEM.m b/src/io/loadConventionalGEM.m index 9fef6715..2154be4f 100644 --- a/src/io/loadConventionalGEM.m +++ b/src/io/loadConventionalGEM.m @@ -34,7 +34,7 @@ % -------- % loadEcModel -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'filename', []; ... 'modelAdapter', []}); filename = p.filename; diff --git a/src/io/loadEcModel.m b/src/io/loadEcModel.m index 9ce5e43f..e6d35d8a 100644 --- a/src/io/loadEcModel.m +++ b/src/io/loadEcModel.m @@ -37,7 +37,7 @@ % -------- % loadConventionalGEM, saveEcModel -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'filename', []; ... 'modelAdapter', []}); filename = p.filename; diff --git a/src/io/loadFluxData.m b/src/io/loadFluxData.m index bd14a404..9ef173a3 100644 --- a/src/io/loadFluxData.m +++ b/src/io/loadFluxData.m @@ -47,7 +47,7 @@ % -------- % loadProtData -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'fluxDataFile', []; ... 'modelAdapter', []}); fluxDataFile = p.fluxDataFile; diff --git a/src/io/loadProtData.m b/src/io/loadProtData.m index 36dee3f9..afe1c376 100644 --- a/src/io/loadProtData.m +++ b/src/io/loadProtData.m @@ -92,7 +92,7 @@ % -------- % loadFluxData -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'protDataFile', []; ... 'filterData', true; ... 'modelAdapter', []; ... diff --git a/src/io/saveEcModel.m b/src/io/saveEcModel.m index 5d96cbc4..94e9ce4d 100644 --- a/src/io/saveEcModel.m +++ b/src/io/saveEcModel.m @@ -45,7 +45,7 @@ function saveEcModel(ecModel,varargin) % -------- % loadEcModel -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'filename', []; ... 'modelAdapter', []}); filename = p.filename; diff --git a/src/kcat_tuning/evotune/bestParsimonious.m b/src/kcat_tuning/evotune/bestParsimonious.m index 8298a655..aac17352 100644 --- a/src/kcat_tuning/evotune/bestParsimonious.m +++ b/src/kcat_tuning/evotune/bestParsimonious.m @@ -25,7 +25,7 @@ % See also % -------- % parsimonyFrontier -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'tolerance', 0.02}); tolerance = p.tolerance; diff --git a/src/kcat_tuning/evotune/cmaesKcatTuning.m b/src/kcat_tuning/evotune/cmaesKcatTuning.m index 48596790..d15c8f98 100644 --- a/src/kcat_tuning/evotune/cmaesKcatTuning.m +++ b/src/kcat_tuning/evotune/cmaesKcatTuning.m @@ -77,7 +77,7 @@ % See also % -------- % screenKcatLeverage, selectTunableMask, tunePriorPenaltyWeight -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... 'evoData', []; ... 'tunableMask', []; ... diff --git a/src/kcat_tuning/evotune/evotuneScore.m b/src/kcat_tuning/evotune/evotuneScore.m index 58049658..dcadc039 100644 --- a/src/kcat_tuning/evotune/evotuneScore.m +++ b/src/kcat_tuning/evotune/evotuneScore.m @@ -54,7 +54,7 @@ % -------- % cmaesKcatTuning, screenKcatLeverage, loadEvotuneData -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... 'maxGrowthWeight', 1.0; ... 'priorPenaltyWeight', 0.0; ... diff --git a/src/kcat_tuning/evotune/isozymeTieMap.m b/src/kcat_tuning/evotune/isozymeTieMap.m index 3075e38a..7675e8c8 100644 --- a/src/kcat_tuning/evotune/isozymeTieMap.m +++ b/src/kcat_tuning/evotune/isozymeTieMap.m @@ -36,7 +36,7 @@ % See also % -------- % cmaesKcatTuning, screenKcatLeverage -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'sources', []; ... 'relTol', 1e-9}); sources = p.sources; diff --git a/src/kcat_tuning/evotune/loadEvotuneData.m b/src/kcat_tuning/evotune/loadEvotuneData.m index 85c300ba..716c6783 100644 --- a/src/kcat_tuning/evotune/loadEvotuneData.m +++ b/src/kcat_tuning/evotune/loadEvotuneData.m @@ -31,7 +31,7 @@ % See also % -------- % cmaesKcatTuning, evotuneScore -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'modelAdapter', []}); modelAdapter = p.modelAdapter; diff --git a/src/kcat_tuning/evotune/nChanged.m b/src/kcat_tuning/evotune/nChanged.m index 23ab967f..0cc93a8c 100644 --- a/src/kcat_tuning/evotune/nChanged.m +++ b/src/kcat_tuning/evotune/nChanged.m @@ -22,7 +22,7 @@ % See also % -------- % tunePriorPenaltyWeight, parsimonyFrontier -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'relTol', 0.02}); relTol = p.relTol; n = sum(abs(log(kcat(:) ./ kcat0(:))) > log1p(relTol)); diff --git a/src/kcat_tuning/evotune/parsimonyFrontier.m b/src/kcat_tuning/evotune/parsimonyFrontier.m index d014f975..fdd9932e 100644 --- a/src/kcat_tuning/evotune/parsimonyFrontier.m +++ b/src/kcat_tuning/evotune/parsimonyFrontier.m @@ -40,7 +40,7 @@ % See also % -------- % bestParsimonious, nChanged -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'thresholds', [0 1 2 3 4 6 8 10 15]}); thresholds = p.thresholds; diff --git a/src/kcat_tuning/evotune/screenKcatLeverage.m b/src/kcat_tuning/evotune/screenKcatLeverage.m index b8862c1c..95592b86 100644 --- a/src/kcat_tuning/evotune/screenKcatLeverage.m +++ b/src/kcat_tuning/evotune/screenKcatLeverage.m @@ -47,7 +47,7 @@ % See also % -------- % selectTunableMask, cmaesKcatTuning -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... 'evoData', []; ... 'tunableMask', []; ... diff --git a/src/kcat_tuning/evotune/selectTunableMask.m b/src/kcat_tuning/evotune/selectTunableMask.m index 50f2eefc..43953cfe 100644 --- a/src/kcat_tuning/evotune/selectTunableMask.m +++ b/src/kcat_tuning/evotune/selectTunableMask.m @@ -30,7 +30,7 @@ % See also % -------- % screenKcatLeverage, cmaesKcatTuning -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'targetImpactShare', 0.9}); targetImpactShare = p.targetImpactShare; diff --git a/src/kcat_tuning/evotune/tunePriorPenaltyWeight.m b/src/kcat_tuning/evotune/tunePriorPenaltyWeight.m index ad4c2c9c..ee4c49f1 100644 --- a/src/kcat_tuning/evotune/tunePriorPenaltyWeight.m +++ b/src/kcat_tuning/evotune/tunePriorPenaltyWeight.m @@ -54,7 +54,7 @@ % See also % -------- % cmaesKcatTuning, screenKcatLeverage, selectTunableMask, nChanged -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... 'evoData', []; ... 'tunableMask', []; ... diff --git a/src/kcat_tuning/sensitivityTuning.m b/src/kcat_tuning/sensitivityTuning.m index 97b397a2..93b4f42f 100644 --- a/src/kcat_tuning/sensitivityTuning.m +++ b/src/kcat_tuning/sensitivityTuning.m @@ -62,7 +62,7 @@ % [model, tunedKcats] = sensitivityTuning(model, 'foldChange', 10); % [model, tunedKcats] = sensitivityTuning(model, desiredGrowthRate, modelAdapter, foldChange, protToIgnore, verbose); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'desiredGrowthRate', []; ... 'modelAdapter', []; ... 'foldChange', 10; ... diff --git a/src/kcat_tuning/sigmaFitter.m b/src/kcat_tuning/sigmaFitter.m index 3eedf138..caadf07e 100644 --- a/src/kcat_tuning/sigmaFitter.m +++ b/src/kcat_tuning/sigmaFitter.m @@ -42,7 +42,7 @@ % [model, sigma] = sigmaFitter(model, 'makePlot', false); % [model, sigma] = sigmaFitter(model, growthRate, Ptot, f, makePlot, modelAdapter); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'growthRate', []; ... 'Ptot', []; ... 'f', []; ... diff --git a/src/limit_proteins/calculateFfactor.m b/src/limit_proteins/calculateFfactor.m index 0230c699..282d674e 100644 --- a/src/limit_proteins/calculateFfactor.m +++ b/src/limit_proteins/calculateFfactor.m @@ -36,7 +36,7 @@ % -------- % loadProtData, fillEnzConcs -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'protData', []; ... 'enzymes', []; ... 'modelAdapter', []}); diff --git a/src/limit_proteins/constrainEnzConcs.m b/src/limit_proteins/constrainEnzConcs.m index 497fd96a..3ae7fca7 100644 --- a/src/limit_proteins/constrainEnzConcs.m +++ b/src/limit_proteins/constrainEnzConcs.m @@ -53,7 +53,7 @@ %Enzyme with numeric entry in model.ec.concs => exchange reaction with %enzyme level as UB -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'removeConstraints', false; ... 'restrictTo', []}); removeConstraints = p.removeConstraints; diff --git a/src/limit_proteins/constrainFluxData.m b/src/limit_proteins/constrainFluxData.m index 26411948..460e3471 100644 --- a/src/limit_proteins/constrainFluxData.m +++ b/src/limit_proteins/constrainFluxData.m @@ -69,7 +69,7 @@ % -------- % loadFluxData -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'fluxData', []; ... 'condition', []; ... 'maxMinGrowth', []; ... diff --git a/src/limit_proteins/fillEnzConcs.m b/src/limit_proteins/fillEnzConcs.m index 341c5042..3faefeb5 100644 --- a/src/limit_proteins/fillEnzConcs.m +++ b/src/limit_proteins/fillEnzConcs.m @@ -54,7 +54,7 @@ % -------- % constrainEnzConcs, loadProtData -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'dataCol', 1}); dataCol = p.dataCol; if isempty(dataCol); dataCol = 1; end diff --git a/src/limit_proteins/flexibilizeEnzConcs.m b/src/limit_proteins/flexibilizeEnzConcs.m index 5b71cc67..f3f8c856 100644 --- a/src/limit_proteins/flexibilizeEnzConcs.m +++ b/src/limit_proteins/flexibilizeEnzConcs.m @@ -96,7 +96,7 @@ % -------- % constrainEnzConcs, getConcControlCoeffs, levelComplexConcs, getComplexFamilies -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'expGrowth', []; ... 'foldChange', []; ... 'iterPerEnzyme', []; ... diff --git a/src/limit_proteins/getComplexFamilies.m b/src/limit_proteins/getComplexFamilies.m index 2c644d0f..a32410e3 100644 --- a/src/limit_proteins/getComplexFamilies.m +++ b/src/limit_proteins/getComplexFamilies.m @@ -59,7 +59,7 @@ % -------- % levelComplexConcs, flexibilizeEnzConcs -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'minOverlap', []; ... 'minShared', []}); minOverlap = p.minOverlap; diff --git a/src/limit_proteins/getConcControlCoeffs.m b/src/limit_proteins/getConcControlCoeffs.m index 9130bb85..01cb1843 100644 --- a/src/limit_proteins/getConcControlCoeffs.m +++ b/src/limit_proteins/getConcControlCoeffs.m @@ -70,7 +70,7 @@ % -------- % flexibilizeEnzConcs -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'proteins', []; ... 'foldChange', []; ... 'limit', []; ... diff --git a/src/limit_proteins/levelComplexConcs.m b/src/limit_proteins/levelComplexConcs.m index b4dfceb7..e5da38af 100644 --- a/src/limit_proteins/levelComplexConcs.m +++ b/src/limit_proteins/levelComplexConcs.m @@ -67,7 +67,7 @@ % -------- % getComplexFamilies, fillEnzConcs, constrainEnzConcs, flexibilizeEnzConcs -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'statistic', []; ... 'fillMissing', []; ... 'families', []}); diff --git a/src/limit_proteins/relaxEnzConcs.m b/src/limit_proteins/relaxEnzConcs.m index bcb72fbc..afe93c0b 100644 --- a/src/limit_proteins/relaxEnzConcs.m +++ b/src/limit_proteins/relaxEnzConcs.m @@ -82,7 +82,7 @@ % -------- % flexibilizeEnzConcs, relaxProteomicsGreedy, constrainEnzConcs, levelComplexConcs -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'expGrowth', []; ... 'bioRxn', []; ... 'perComplex', false; ... diff --git a/src/limit_proteins/relaxProteomicsGreedy.m b/src/limit_proteins/relaxProteomicsGreedy.m index 1f878805..75202f5b 100644 --- a/src/limit_proteins/relaxProteomicsGreedy.m +++ b/src/limit_proteins/relaxProteomicsGreedy.m @@ -67,7 +67,7 @@ % -------- % flexibilizeEnzConcs, getEnzymeBottlenecks -p = parseGECKOargs(varargin, {'minimalGrowth', []; 'enzymeSet', []; ... +p = parseRAVENargs(varargin, {'minimalGrowth', []; 'enzymeSet', []; ... 'maxIterations', []; 'defaultUpperBound', []}); minimalGrowth = p.minimalGrowth; enzymeSet = p.enzymeSet; diff --git a/src/limit_proteins/zeroUnmeasuredIsozymes.m b/src/limit_proteins/zeroUnmeasuredIsozymes.m index be186e6c..e1f004e0 100644 --- a/src/limit_proteins/zeroUnmeasuredIsozymes.m +++ b/src/limit_proteins/zeroUnmeasuredIsozymes.m @@ -58,7 +58,7 @@ % -------- % getIsozymeCoverage, constrainEnzConcs, fillEnzConcs -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'method', 'zero'; ... 'limit', []}); method = p.method; diff --git a/src/utilities/createGECKOproject.m b/src/utilities/createGECKOproject.m index 94a84244..c55d83c2 100644 --- a/src/utilities/createGECKOproject.m +++ b/src/utilities/createGECKOproject.m @@ -25,7 +25,7 @@ function createGECKOproject(varargin) % createGECKOproject(name, path); % createGECKOproject('name', name); -p = parseGECKOargs(varargin, { ... +p = parseRAVENargs(varargin, { ... 'name', []; ... 'path', []; ... 'adapterFormat', 'toml'}); diff --git a/src/utilities/parseGECKOargs.m b/src/utilities/parseGECKOargs.m deleted file mode 100644 index bcb5fec9..00000000 --- a/src/utilities/parseGECKOargs.m +++ /dev/null @@ -1,98 +0,0 @@ -function args = parseGECKOargs(rawArgs, spec) -% parseGECKOargs Resolve optional arguments passed positionally or by name. -% -% Helper that lets a GECKO function accept its optional arguments either in -% the traditional positional order or as name-value pairs, without losing -% backward compatibility. The required leading arguments stay explicit in the -% function signature; the optional tail is collected into varargin and passed -% here together with a specification of the optional parameters. -% -% Parameters -% ---------- -% rawArgs : cell -% the function's varargin, i.e. whatever was supplied after the required -% positional arguments. -% spec : cell -% an N-by-2 or N-by-3 cell array describing the optional parameters, one -% per row: -% -% - column 1 : parameter name (char). -% - column 2 : default value, used when the parameter is not supplied. -% - column 3 : optional validator, a function handle that is called with -% the resolved value and should error on an invalid value. Use [] for -% no validation. -% -% Returns -% ------- -% args : struct -% a struct with one field per parameter name, holding the resolved value. -% -% Notes -% ----- -% The calling convention is detected from rawArgs: it is treated as name- -% value pairs only when it has an even number of elements AND every element -% in a name position is a string matching one of the declared parameter -% names. Otherwise it is treated as positional and assigned in the order of -% the spec. A call therefore uses either the positional or the name-value -% form for the optional tail, not a mix of the two. When an optional tail -% value could itself be a string equal to a parameter name, use the explicit -% name-value form to remove the ambiguity. -% -% Examples -% -------- -% % inside makeEcModel(model, varargin): -% p = parseGECKOargs(varargin, { ... -% 'geckoLight', false; ... -% 'modelAdapter', []}); -% geckoLight = p.geckoLight; -% modelAdapter = p.modelAdapter; -% -% % the caller may then use either form: -% makeEcModel(model, true, adapter) % positional -% makeEcModel(model, 'geckoLight', true) % named - -if nargin < 2 || isempty(spec) - error('parseGECKOargs:noSpec', 'A parameter specification is required.'); -end - -names = spec(:, 1); -defaults = spec(:, 2); -hasValidators = size(spec, 2) >= 3; - -% Seed every parameter with its default. -args = struct(); -for i = 1:numel(names) - args.(names{i}) = defaults{i}; -end - -if ~isempty(rawArgs) - isNameValue = mod(numel(rawArgs), 2) == 0 && ... - all(cellfun(@(x) (ischar(x) || isstring(x)) && isscalar(string(x)) && ... - any(strcmpi(x, names)), rawArgs(1:2:end))); - - if isNameValue - for k = 1:2:numel(rawArgs) - args.(names{strcmpi(rawArgs{k}, names)}) = rawArgs{k + 1}; - end - else - if numel(rawArgs) > numel(names) - error('parseGECKOargs:tooManyArgs', ... - 'Too many positional arguments: expected at most %d, got %d.', ... - numel(names), numel(rawArgs)); - end - for k = 1:numel(rawArgs) - args.(names{k}) = rawArgs{k}; - end - end -end - -% Apply optional validators. -if hasValidators - for i = 1:numel(names) - validator = spec{i, 3}; - if ~isempty(validator) - validator(args.(names{i})); - end - end -end -end diff --git a/src/utilities/upgradeGECKOproject.m b/src/utilities/upgradeGECKOproject.m index 092c7f86..1630c999 100644 --- a/src/utilities/upgradeGECKOproject.m +++ b/src/utilities/upgradeGECKOproject.m @@ -53,7 +53,7 @@ % convertAdapterToToml, convertCustomKcats, upgradeEcModel, createGECKOproject checkOptions(varargin, {'dryRun','ecModel'}, 'upgradeGECKOproject'); -p = parseGECKOargs(varargin, {'dryRun', true; 'ecModel', []}); +p = parseRAVENargs(varargin, {'dryRun', true; 'ecModel', []}); projectPath = char(projectPath); if ~isfolder(projectPath) error('upgradeGECKOproject:path', 'No folder %s.', projectPath) diff --git a/test/unit_tests/geckoCoreFunctionTests.m b/test/unit_tests/geckoCoreFunctionTests.m index 5ed09662..9b92ea06 100644 --- a/test/unit_tests/geckoCoreFunctionTests.m +++ b/test/unit_tests/geckoCoreFunctionTests.m @@ -337,11 +337,10 @@ function testfuzzyKcatMatchingMedian_tc0060(testCase) 'modelAdapter', adapter, 'kcatAggregation', 'mean'), ... 'fuzzyKcatMatching:invalidKcatAggregation') - % Mixing positional and name-value arguments puts a name where a - % value belongs; that has to fail rather than quietly wildcard every - % EC number, which is what an unvalidated forceWClvl would do. - verifyError(testCase, @() fuzzyKcatMatching(ecModel, [], adapter, ... - 'kcatAggregation', 'median'), 'fuzzyKcatMatching:invalidForceWClvl') + % Leading positional arguments may be followed by name-value pairs. + kcatListMixed = fuzzyKcatMatching(ecModel, [], adapter, ... + 'kcatAggregation', 'median'); + verifyEqual(testCase, kcatListMixed.kcats, kcatListMedian.kcats) end function testfuzzyKcatMatching_tc0010(testCase) From f84240c1be65cf8e24c9b64574e5b1a6f5ed909f Mon Sep 17 00:00:00 2001 From: Eduard Kerkhoven Date: Fri, 20 Nov 2026 21:34:58 +0100 Subject: [PATCH 2/2] refactor: score and constrain conditions with the experiments functions of RAVEN constrainFluxData, loadEvotuneData and evotuneScore call RAVEN's applyFluxDataConstraints, loadExperimentData, simulateConditions, computeCarbonWeights and experimentDistance, and add the biomass reaction, the carbon source, the anaerobic and protein-content functions and the kcat penalty of the model adapter. loadFluxData and fillCarbonNum are removed; RAVEN has loadFluxData(path) and computeCarbonWeights. The experimental data files are fluxData.tsv, maxGrowth.tsv and zeroFlux.tsv, with the column weight, and the data structure has the fields fluxData, maxGrowth and zeroFlux. The evotune functions take the argument experimentData (was evoData). evotuneScore returns the carbon weights as its fourth output and takes them as the argument carbonWeights, so they are computed once per run. addCarbonNum and getcarbonnum raise GECKO:removedFunction and name computeCarbonWeights. GECKOInstaller checks that RAVEN has the experiments functions. --- GECKOInstaller.m | 16 + src/deprecated/addCarbonNum.m | 16 +- src/deprecated/bayesianSensitivityTuning.m | 4 +- src/deprecated/getcarbonnum.m | 6 +- src/io/loadFluxData.m | 155 --------- src/kcat_tuning/evotune/cmaesKcatTuning.m | 45 +-- src/kcat_tuning/evotune/evotuneScore.m | 158 +++------ src/kcat_tuning/evotune/loadEvotuneData.m | 56 +-- src/kcat_tuning/evotune/parsimonyFrontier.m | 4 +- src/kcat_tuning/evotune/screenKcatLeverage.m | 41 +-- .../evotune/tunePriorPenaltyWeight.m | 14 +- src/kcat_tuning/fillCarbonNum.m | 327 ------------------ src/limit_proteins/constrainFluxData.m | 99 +----- src/utilities/upgradeGECKOproject.m | 2 +- test/VerificationMatrix.tsv | 5 +- test/unit_tests/geckoCoreFunctionTests.m | 19 +- test/unit_tests/parityRound5G3Tests.m | 40 +-- test/unit_tests/parityRound5G4Tests.m | 48 ++- .../full_ecModel/data/evotuneFluxData.tsv | 34 -- .../full_ecModel/data/evotuneMaxGrowth.tsv | 9 - tutorials/full_ecModel/data/fluxData.tsv | 37 +- tutorials/full_ecModel/data/maxGrowth.tsv | 9 + .../{evotuneZeroExch.tsv => zeroFlux.tsv} | 0 tutorials/full_ecModel/protocol.m | 6 +- 24 files changed, 258 insertions(+), 892 deletions(-) delete mode 100644 src/io/loadFluxData.m delete mode 100644 src/kcat_tuning/fillCarbonNum.m delete mode 100644 tutorials/full_ecModel/data/evotuneFluxData.tsv delete mode 100644 tutorials/full_ecModel/data/evotuneMaxGrowth.tsv create mode 100644 tutorials/full_ecModel/data/maxGrowth.tsv rename tutorials/full_ecModel/data/{evotuneZeroExch.tsv => zeroFlux.tsv} (100%) diff --git a/GECKOInstaller.m b/GECKOInstaller.m index 55f85aff..e6af21e1 100644 --- a/GECKOInstaller.m +++ b/GECKOInstaller.m @@ -11,6 +11,7 @@ paths = GECKOInstaller.GetFilteredSubPaths(sourceDir, GECKOInstaller.FILE_FILTER); GECKOInstaller.checkRAVENversion('3.0.0'); % Minimum RAVEN version + GECKOInstaller.checkRAVENExperiments(); % loadExperimentData and related functions of RAVEN % Check unique function names if ~exist("checkFunctionUniqueness.m") @@ -61,6 +62,21 @@ newPaths = char(join(pathsLeft, pathSep)); end + function checkRAVENExperiments() + % GECKO scores and constrains models against measured conditions with + % the functions of the experiments folder of RAVEN (loadFluxData, + % loadExperimentData, simulateConditions, computeCarbonWeights, + % experimentDistance, applyFluxDataConstraints). The version check + % ignores the pre-release suffix of RAVEN, so it cannot tell a RAVEN + % that has these functions from one that has not. + if exist('loadExperimentData','file') ~= 2 || exist('applyFluxDataConstraints','file') ~= 2 + error(['The installed RAVEN does not have the experiments functions ' ... + '(loadExperimentData, applyFluxDataConstraints, ...), which GECKO requires. ' ... + 'Upgrade RAVEN by following the instructions available ' ... + 'here.']) + end + end + function checkRAVENversion(minmVer) wrongVersion = false; try diff --git a/src/deprecated/addCarbonNum.m b/src/deprecated/addCarbonNum.m index 1764b346..4a3acf91 100644 --- a/src/deprecated/addCarbonNum.m +++ b/src/deprecated/addCarbonNum.m @@ -1,14 +1,14 @@ -function varargout = addCarbonNum(varargin) -% addCarbonNum Deprecated alias for fillCarbonNum. +function varargout = addCarbonNum(varargin) %#ok +% addCarbonNum Removed: carbon numbers of exchange reactions of GECKO 3. % -% Kept for backward compatibility with pre-GECKO4 code. Will be removed -% in a future release; switch to fillCarbonNum. +% This function raises an error that names its replacement, RAVEN's +% computeCarbonWeights, which returns the carbon weight of each reaction as a +% containers.Map instead of writing the field model.excarbon. % % See also % -------- -% fillCarbonNum +% computeCarbonWeights -warning('GECKO:deprecatedName', ... - 'addCarbonNum is deprecated; use fillCarbonNum instead. The old name will be removed in a future release.'); -[varargout{1:nargout}] = fillCarbonNum(varargin{:}); +error('GECKO:removedFunction', ['addCarbonNum was removed in GECKO 4. Use computeCarbonWeights (RAVEN),' ... + ' which returns the carbon weight of each reaction as a containers.Map.']); end diff --git a/src/deprecated/bayesianSensitivityTuning.m b/src/deprecated/bayesianSensitivityTuning.m index 9275a5e9..75261cc7 100644 --- a/src/deprecated/bayesianSensitivityTuning.m +++ b/src/deprecated/bayesianSensitivityTuning.m @@ -3,8 +3,8 @@ % % GECKO 4 tunes kcats against experimental data with cmaesKcatTuning (the % advanced kcat tuning), which reads the same kinds of data from -% data/evotuneMaxGrowth.tsv, data/evotuneFluxData.tsv and -% data/evotuneZeroExch.tsv (see loadEvotuneData). sensitivityTuning is the +% data/maxGrowth.tsv, data/fluxData.tsv and +% data/zeroFlux.tsv (see loadEvotuneData). sensitivityTuning is the % standard kcat tuning. This function raises an error that names them. % % See also diff --git a/src/deprecated/getcarbonnum.m b/src/deprecated/getcarbonnum.m index 798bb6a9..735ab31b 100644 --- a/src/deprecated/getcarbonnum.m +++ b/src/deprecated/getcarbonnum.m @@ -5,8 +5,8 @@ % % See also % -------- -% fillCarbonNum +% computeCarbonWeights -error('GECKO:removedFunction', ['getcarbonnum was removed in GECKO 4. Use fillCarbonNum, which writes' ... - ' the carbon number of each exchange reaction to model.excarbon.']); +error('GECKO:removedFunction', ['getcarbonnum was removed in GECKO 4. Use computeCarbonWeights (RAVEN),' ... + ' which returns the carbon weight of each reaction as a containers.Map.']); end diff --git a/src/io/loadFluxData.m b/src/io/loadFluxData.m deleted file mode 100644 index 9ef173a3..00000000 --- a/src/io/loadFluxData.m +++ /dev/null @@ -1,155 +0,0 @@ -function fluxData = loadFluxData(varargin) -% loadFluxData Load total protein measurements and exchange flux data. -% -% Loads total protein content and exchange flux data (e.g. carbon source, -% O2, CO2) from a tab-delimited file. -% -% Name-Value Arguments -% -------------------- -% fluxDataFile : char -% path to file with flux data (default reads data/fluxData.tsv as -% specified in modelAdapter). -% modelAdapter : ModelAdapter -% a loaded model adapter (default: the current default model adapter). -% -% Returns -% ------- -% fluxData : struct -% structure with flux data. -% -% Notes -% ----- -% The function raises an error when the file does not exist -% (loadFluxData:fileNotFound), when a line has a different number of cells -% than the header (loadFluxData:raggedRow) or when the file has no exchange -% flux columns (loadFluxData:noExchangeColumns). An exchange column header has -% the form ' ()'; the reaction id may contain -% parentheses. -% -% The fluxData structure has the following fields: -% -% - conds : sampling condition. -% - Ptot : total protein (g/gDCW). -% - grRate : growth rate (1/h). -% - exchFluxes : exchange fluxes (mmol/gDCWh). -% - exchMets : exchanged metabolites, matching exchFluxes. -% - exchRxnIDs : exchange reaction IDs, matching exchMets. -% - evotuneRMSEweight : if column existed in fluxData.tsv, weights for RMSE calculation in evotune kcat tuning. -% - source : if column existed in fluxData.tsv, description where the data comes from. -% -% Examples -% -------- -% % optional arguments may be given positionally or as name-value pairs: -% fluxData = loadFluxData(fluxDataFile, modelAdapter); -% fluxData = loadFluxData('fluxDataFile', fluxDataFile); -% -% See also -% -------- -% loadProtData - -p = parseRAVENargs(varargin, { ... - 'fluxDataFile', []; ... - 'modelAdapter', []}); -fluxDataFile = p.fluxDataFile; -modelAdapter = p.modelAdapter; - -if isempty(modelAdapter) - modelAdapter = ModelAdapterManager.getDefault(); - if isempty(modelAdapter) - error('Either send in a modelAdapter or set the default model adapter in the ModelAdapterManager.') - end -end -params = modelAdapter.getParameters(); - -if isempty(fluxDataFile) - fluxDataFile = fullfile(params.path,'data','fluxData.tsv'); -end - -%Load total protein content and flux data -if ~isfile(fluxDataFile) - error('loadFluxData:fileNotFound', 'The flux data file %s does not exist.', fluxDataFile); -end -fID = fopen(fluxDataFile); -formatStr = '%s'; -data = textscan(fID,formatStr,'Delimiter','\n'); -fclose(fID); -fluxDataRaw = []; -for i=1:length(data{1}) - row = data{1}(i); - row = strsplit(row{1},'\t'); - row = row(1:end); - if i > 1 && numel(row) ~= size(fluxDataRaw,2) - error('loadFluxData:raggedRow', ... - 'Line %d of %s has %d cells, the header has %d.', ... - i, fluxDataFile, numel(row), size(fluxDataRaw,2)); - end - fluxDataRaw = [fluxDataRaw; row]; -end -if size(fluxDataRaw,2) == 0 - error('The input file %s is likely not tab-delimited.', fluxDataFile) -end -fluxData = []; - -%Find additional fields -extraFieldNames = {'evotuneRMSEweight','source'}; -extraFieldNumeric = [true,false]; -[logicalExtra, extraFields] = ismember(extraFieldNames,fluxDataRaw(1,:)); -extraFields(~logicalExtra) = []; -extraFieldNames(~logicalExtra) = []; -extraFieldNumeric(~logicalExtra) = []; -if any(extraFields) - for i=1:numel(extraFields) - if extraFieldNumeric(i) - fluxData.(extraFieldNames{i}) = str2double(fluxDataRaw(2:end,extraFields(i))); - else - fluxData.(extraFieldNames{i}) = fluxDataRaw(2:end,extraFields(i)); - end - end - fluxDataRaw(:,extraFields) = []; -end - -%Extract observed byProduct names from file -if size(fluxDataRaw,2) < 4 - error('loadFluxData:noExchangeColumns', ... - '%s has no exchange flux columns after the columns condition, Ptot and grRate.', fluxDataFile); -end -exchRxns = fluxDataRaw(1,4:end); -exchMets = exchRxns; -for i=1:numel(exchRxns) - [exchMets{i}, exchRxns{i}] = splitExchHeader(exchRxns{i}); -end - -fluxData.conds = fluxDataRaw(2:end,1); -fluxData.Ptot = str2double(fluxDataRaw(2:end,2)); -fluxData.grRate = str2double(fluxDataRaw(2:end,3)); -fluxData.exchFluxes = str2double(fluxDataRaw(2:end,4:end)); -%fluxData.exchFluxes(isnan(fluxData.exchFluxes)) = 0; -fluxData.exchMets = exchMets; -fluxData.exchRxnIDs = exchRxns; -end - -function [met, rxn] = splitExchHeader(header) -% Split a column header ' ()' at the parenthesis that -% matches the final closing one, so a reaction id may contain parentheses. A -% header of another form gives the whole header as the reaction id. -met = strtrim(header); -rxn = header; -if ~endsWith(header,')') - return -end -depth = 0; -for k=numel(header):-1:1 - if header(k) == ')' - depth = depth + 1; - elseif header(k) == '(' - depth = depth - 1; - if depth == 0 - if k < numel(header)-1 - met = strtrim(header(1:k-1)); - rxn = header(k+1:end-1); - end - return - end - end -end -end diff --git a/src/kcat_tuning/evotune/cmaesKcatTuning.m b/src/kcat_tuning/evotune/cmaesKcatTuning.m index d15c8f98..fc7f1426 100644 --- a/src/kcat_tuning/evotune/cmaesKcatTuning.m +++ b/src/kcat_tuning/evotune/cmaesKcatTuning.m @@ -29,9 +29,9 @@ % -------------------- % modelAdapter : ModelAdapter % a loaded model adapter (default: the current default model adapter). -% evoData : struct +% experimentData : struct % as returned by loadEvotuneData (default: loaded from modelAdapter). -% Without fluxData and maxGrate, the error evotune:noData is raised. +% Without fluxData and maxGrowth, the error evotune:noData is raised. % tunableMask, screen, targetImpactShare % control which kcats are searched over, in order of precedence. Pass % tunableMask to fix the set yourself. Otherwise, a mask is built by @@ -79,7 +79,7 @@ % screenKcatLeverage, selectTunableMask, tunePriorPenaltyWeight p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... - 'evoData', []; ... + 'experimentData', []; ... 'tunableMask', []; ... 'screen', []; ... 'targetImpactShare', 0.9; ... @@ -88,7 +88,7 @@ 'seed', []; ... 'verbose', true}); modelAdapter = p.modelAdapter; -evoData = p.evoData; +experimentData = p.experimentData; tunableMask = p.tunableMask; screen = p.screen; targetImpactShare = p.targetImpactShare; @@ -103,19 +103,19 @@ error('Either send in a modelAdapter or set the default ecModel adapter in the ModelAdapterManager.') end end -if isempty(evoData) - evoData = loadEvotuneData('modelAdapter', modelAdapter); +if isempty(experimentData) + experimentData = loadEvotuneData('modelAdapter', modelAdapter); end -if isempty(evoData.fluxData) && isempty(evoData.maxGrate) - error('evotune:noData', ['evoData has neither fluxData nor maxGrate (from ' ... - 'data/evotuneFluxData.tsv and data/evotuneMaxGrowth.tsv); there is nothing to tune against.']) +if isempty(experimentData.fluxData) && isempty(experimentData.maxGrowth) + error('evotune:noData', ['experimentData has neither fluxData nor maxGrowth (from ' ... + 'data/fluxData.tsv and data/maxGrowth.tsv); there is nothing to tune against.']) end params = modelAdapter.params.evotune; if isempty(tunableMask) if isempty(screen) screen = screenKcatLeverage(ecModel, 'modelAdapter',modelAdapter, ... - 'evoData',evoData, 'nProc',nProc); + 'experimentData',experimentData, 'nProc',nProc); end tunableMask = selectTunableMask(ecModel, screen, 'targetImpactShare',targetImpactShare); end @@ -143,10 +143,10 @@ expand = @(Xlog) exp(Xlog(assign,:)); -% Baseline: score the untouched prior, and keep the excarbon-augmented -% model to use in the rest of the run (see evotuneScore's 4th -% output -- avoids rescanning metabolite formulas on every candidate). -[obj0,rmse0,~,baseModel] = scoreCandidate(ecModel,tunableIdx,ecRxnIdsTunable,evoData,modelAdapter,params,kcat0,kcat0,sigma0log); +% Baseline: score the untouched prior, and keep the carbon weights to use in +% the rest of the run (see evotuneScore's 4th output -- avoids rescanning +% metabolite formulas on every candidate). +[obj0,rmse0,~,carbonWeights] = scoreCandidate(ecModel,tunableIdx,ecRxnIdsTunable,experimentData,modelAdapter,params,kcat0,kcat0,sigma0log,[]); bestObj = obj0; bestRmse = rmse0; bestVec = kcat0; rmseTrace = []; objectiveTrace = []; @@ -169,15 +169,15 @@ objVals = zeros(popsize,1); rmseVals = zeros(popsize,1); if nProc == 1 for k = 1:popsize - [objVals(k),rmseVals(k),~,baseModel] = scoreCandidate( ... - baseModel,tunableIdx,ecRxnIdsTunable,evoData,modelAdapter,params, ... - kcatMatrix(:,k),kcat0,sigma0log); + [objVals(k),rmseVals(k)] = scoreCandidate( ... + ecModel,tunableIdx,ecRxnIdsTunable,experimentData,modelAdapter,params, ... + kcatMatrix(:,k),kcat0,sigma0log,carbonWeights); end else parfor k = 1:popsize [objVals(k),rmseVals(k)] = scoreCandidate( ... - ecModel,tunableIdx,ecRxnIdsTunable,evoData,modelAdapter,params, ... - kcatMatrix(:,k),kcat0,sigma0log); + ecModel,tunableIdx,ecRxnIdsTunable,experimentData,modelAdapter,params, ... + kcatMatrix(:,k),kcat0,sigma0log,carbonWeights); end end generation = generation + 1; @@ -234,13 +234,14 @@ %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Scoring core, shared with screenKcatLeverage/tunePriorPenaltyWeight. -function [obj,rmse,rmseDetail,ecModel] = scoreCandidate(ecModel,tunableIdx,ecRxnIdsTunable,evoData,modelAdapter,params,kcatVec,kcat0,sigma0log) +function [obj,rmse,rmseDetail,carbonWeights] = scoreCandidate(ecModel,tunableIdx,ecRxnIdsTunable,experimentData,modelAdapter,params,kcatVec,kcat0,sigma0log,carbonWeights) ecModel.ec.kcat(tunableIdx) = kcatVec; ecModel = applyKcatConstraints(ecModel,'updateRxns',ecRxnIdsTunable); -[obj,rmse,rmseDetail,ecModel] = evotuneScore(ecModel,evoData, ... +[obj,rmse,rmseDetail,carbonWeights] = evotuneScore(ecModel,experimentData, ... 'modelAdapter',modelAdapter, 'maxGrowthWeight',params.maxGrowthWeight, ... 'priorPenaltyWeight',params.priorPenaltyWeight, ... - 'kcatVec',kcatVec, 'kcat0',kcat0, 'sigma0log',sigma0log); + 'kcatVec',kcatVec, 'kcat0',kcat0, 'sigma0log',sigma0log, ... + 'carbonWeights',carbonWeights); end function [lo,hi] = kcatBoundsLocal(kcat0) diff --git a/src/kcat_tuning/evotune/evotuneScore.m b/src/kcat_tuning/evotune/evotuneScore.m index dcadc039..0d26aefd 100644 --- a/src/kcat_tuning/evotune/evotuneScore.m +++ b/src/kcat_tuning/evotune/evotuneScore.m @@ -1,18 +1,21 @@ -function [objective,rmse,rmseDetail,ecModel] = evotuneScore(ecModel,evoData,varargin) -% evotuneScore Score one ecModel's current kcats against evotune data. +function [objective,rmse,rmseDetail,carbonWeights] = evotuneScore(ecModel,experimentData,varargin) +% evotuneScore Score one ecModel's current kcats against the experimental data. % % Internal helper used by cmaesKcatTuning, screenKcatLeverage and % tunePriorPenaltyWeight to score one ecModel (with a given set of kcats % already applied via applyKcatConstraints) against the experimental -% growth and flux data. Fuses what geckopy splits into simulate.py (the -% FBA half) and distance.py (the RMSE half) into one function. +% growth and flux data. The conditions are simulated with RAVEN's +% simulateConditions and scored with experimentDistance; this function +% supplies the model-specific parts: the biomass reaction, the anaerobic +% and protein-content functions of the model adapter, and the penalty on +% the kcats. % % Parameters % ---------- % ecModel : struct % ecModel with the current candidate kcats already applied (via % applyKcatConstraints). Not mutated. -% evoData : struct +% experimentData : struct % structure with experimental data to be used, as returned by % loadEvotuneData. % @@ -22,7 +25,7 @@ % a loaded model adapter (default: the current default model adapter). % maxGrowthWeight : double % relative weight of the max-growth dataset against the flux dataset: -% (rmseFlux + w*rmseMaxGrate) / (w+1) (default 1.0). +% (rmseFlux + w*rmseMaxGrowth) / (w+1) (default 1.0). % priorPenaltyWeight : double % weight on a prior term added to the objective (default 0, i.e. the % objective equals the plain RMSE). See kcatVec/kcat0/sigma0log below. @@ -31,6 +34,9 @@ % current values, prior values, and per-kcat prior log-space std dev, % all the same length and in the same order. The penalty added to the % objective is priorPenaltyWeight * mean((log(kcatVec./kcat0)./sigma0log).^2). +% carbonWeights : containers.Map +% carbon weights of the reactions, as computed by computeCarbonWeights +% (default: computed here). % % Returns % ------- @@ -42,17 +48,16 @@ % regardless of priorPenaltyWeight. % rmseDetail : cell % per-condition RMSE values, labelled by data type and condition index. -% ecModel : struct -% the input ecModel, with ec.excarbon cached on it if it wasn't -% already present. MATLAB structs are copied on assignment, not -% shared by reference like a Python object, so callers scoring many -% candidates in a loop should pass this output back in as the base -% copy for the next candidate, rather than recomputing excarbon (an -% expensive metabolite-formula scan) on every single call. +% carbonWeights : containers.Map +% the carbon weights that were used. The weights depend on the +% metabolite formulas and not on the kcats, so callers scoring many +% candidates in a loop should pass this output back in as carbonWeights +% for the next candidate, rather than recomputing it (an expensive +% metabolite-formula scan) on every call. % % See also % -------- -% cmaesKcatTuning, screenKcatLeverage, loadEvotuneData +% cmaesKcatTuning, screenKcatLeverage, loadEvotuneData, experimentDistance p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... @@ -60,10 +65,12 @@ 'priorPenaltyWeight', 0.0; ... 'kcatVec', []; ... 'kcat0', []; ... - 'sigma0log', []}); + 'sigma0log', []; ... + 'carbonWeights', []}); modelAdapter = p.modelAdapter; maxGrowthWeight = p.maxGrowthWeight; priorPenaltyWeight = p.priorPenaltyWeight; +carbonWeights = p.carbonWeights; if isempty(modelAdapter) modelAdapter = ModelAdapterManager.getDefault(); @@ -71,36 +78,44 @@ error('Either send in a modelAdapter or set the default ecModel adapter in the ModelAdapterManager.') end end +bioRxnId = modelAdapter.params.bioRxn; -if ~isfield(ecModel,'excarbon') - ecModel = fillCarbonNum(ecModel); - ecModel.excarbon(ecModel.excarbon == 0) = 1; +if isempty(carbonWeights) + rxnIds = experimentData.zeroFlux(:); + if ~isempty(experimentData.fluxData) + rxnIds = [rxnIds; experimentData.fluxData.exchRxnIDs(:)]; + end + if ~isempty(experimentData.maxGrowth) + rxnIds = [rxnIds; experimentData.maxGrowth.exchRxnIDs(:)]; + end + carbonWeights = computeCarbonWeights(ecModel, unique([rxnIds; {bioRxnId}]), bioRxnId); end -rmse_1 = []; rmseDetail1 = []; rmse_2 = []; rmseDetail2 = []; +makeAnaerobic = @(model) modelAdapter.makeModelAnaerobic(model); +changeProteinBiomass = @(model,ptot) modelAdapter.changeProteinBiomass(model,ptot); -if ~isempty(evoData.fluxData) - [rmse_1, rmseDetail1] = simulateCondition(ecModel,evoData.fluxData,true,evoData.zeroFlux,modelAdapter); - rmseDetail1 = [cellstr("fluxData_" + string(1:numel(rmseDetail1)))',num2cell(rmseDetail1)]; +fluxSims = []; +maxGrowthSims = []; +if ~isempty(experimentData.fluxData) + fluxSims = simulateConditions(ecModel, experimentData.fluxData, true, ... + experimentData.zeroFlux, bioRxnId, 'makeAnaerobic', makeAnaerobic, ... + 'changeProteinBiomass', changeProteinBiomass); end -if ~isempty(evoData.maxGrate) - [rmse_2, rmseDetail2] = simulateCondition(ecModel,evoData.maxGrate,false,evoData.zeroFlux,modelAdapter); - rmseDetail2 = [cellstr("maxGrowth_" + string(1:numel(rmseDetail2)))',num2cell(rmseDetail2)]; +if ~isempty(experimentData.maxGrowth) + maxGrowthSims = simulateConditions(ecModel, experimentData.maxGrowth, false, ... + {}, bioRxnId, 'makeAnaerobic', makeAnaerobic, ... + 'changeProteinBiomass', changeProteinBiomass); end +[rmse, detail] = experimentDistance(experimentData, fluxSims, maxGrowthSims, ... + carbonWeights, bioRxnId, 'maxGrowthWeight', maxGrowthWeight); -parts = []; weights = []; -if ~isempty(rmse_1) - parts(end+1) = rmse_1; weights(end+1) = 1.0; +rmseDetail = []; +if isfield(detail,'fluxData') + rmseDetail = [rmseDetail; [cellstr("fluxData_" + string(1:numel(detail.fluxData)))', num2cell(detail.fluxData(:))]]; end -if ~isempty(rmse_2) - parts(end+1) = rmse_2; weights(end+1) = maxGrowthWeight; +if isfield(detail,'maxGrowth') + rmseDetail = [rmseDetail; [cellstr("maxGrowth_" + string(1:numel(detail.maxGrowth)))', num2cell(detail.maxGrowth(:))]]; end -if isempty(parts) - rmse = NaN; -else - rmse = sum(parts .* weights) / sum(weights); -end -rmseDetail = [rmseDetail1;rmseDetail2]; objective = rmse; if priorPenaltyWeight > 0 && ~isempty(p.kcatVec) && ~isempty(p.kcat0) && ~isempty(p.sigma0log) @@ -108,74 +123,3 @@ objective = rmse + priorPenaltyWeight * mean(dev.^2); end end - -%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% -function [rmse, rmseList] = simulateCondition(ecModel,data,constrain,rxn2block,modelAdapter) -% One condition's FBA simulation + carbon-weighted RMSE against measured -% data. - -rmseList = zeros(length(data.conds),1); - -[~,exchIdx] = ismember(data.conds,data.exchMets); -if any(exchIdx == 0) - missingMets = strjoin(unique(data.conds(exchIdx == 0)), '; '); - error('Carbon source(s) "%s" in the provided fluxData or maxGrowth cannot be matched by name with an exchange reaction.', missingMets); -end - -for i = 1:length(data.conds) - % Set all other carbon sources to zero, then unblock this condition's - model_tmp = setParam(ecModel,'lb',data.exchRxnIDs(unique(exchIdx)),0); - if constrain % RMSE from flux data: constrain carbon uptake at the measured rate - model_tmp = setParam(model_tmp,'lb',data.exchRxnIDs(exchIdx(i)),data.exchFluxes(i,exchIdx(i))); - else % RMSE from max growth: leave carbon uptake fully open - model_tmp = setParam(model_tmp,'lb',data.exchRxnIDs(exchIdx(i)),-1000); - end - - % This row's own carbon source is never itself asserted zero-flux. - rxn2blockIter = rxn2block; - rxn2blockIter(strcmp(rxn2blockIter,data.exchRxnIDs(exchIdx(i)))) = []; - - o2Flux = data.exchFluxes(i,strcmp('oxygen',data.exchMets)); - if o2Flux == 0 - model_tmp = modelAdapter.makeModelAnaerobic(model_tmp); - end - if ~(isnan(data.Ptot(i)) || data.Ptot(i) == 0) - model_tmp = modelAdapter.changeProteinBiomass(model_tmp,data.Ptot(i)); - end - - sol = solveLP(model_tmp); - if checkSolution(sol) - bioRxn = getIndexes(model_tmp,data.biomass,'rxns'); - % Biomass has an assumed 41 Cmmol/gDCW, matching fillCarbonNum. - bioMeas = data.grRate(i) * 41; - bioSim = sol.x(bioRxn) * 41; - - if constrain - fluxToCheck = ~isnan(data.exchFluxes(i,:)); - fluxRxnIdx = getIndexes(model_tmp,data.exchRxnIDs(fluxToCheck),'rxns'); - cNormMeasFlux = model_tmp.excarbon(fluxRxnIdx) .* transpose(data.exchFluxes(i,fluxToCheck)); - cNormSimFlux = model_tmp.excarbon(fluxRxnIdx) .* sol.x(fluxRxnIdx); - - blockRxnIdx = getIndexes(model_tmp,rxn2blockIter,'rxns'); - blockSim = sol.x(blockRxnIdx) .* model_tmp.excarbon(blockRxnIdx); - % Already-correct zero predictions don't dilute the RMSE. - blockSim(blockSim == 0) = []; - blockMeas = zeros(numel(blockSim),1); - - measured = [cNormMeasFlux; bioMeas; blockMeas]; - simulated = [cNormSimFlux; bioSim; blockSim]; - else - measured = bioMeas; - simulated = bioSim; - end - rmseList(i) = sqrt(mean((measured - simulated).^2)); - else - rmseList(i) = NaN; - end -end -rmseList(isnan(rmseList)) = 99; % Fixed penalty for an infeasible condition. -if isfield(data,'evotuneRMSEweight') && ~isempty(data.evotuneRMSEweight) - rmseList = rmseList .* data.evotuneRMSEweight; -end -rmse = mean(rmseList); -end diff --git a/src/kcat_tuning/evotune/loadEvotuneData.m b/src/kcat_tuning/evotune/loadEvotuneData.m index 716c6783..d0677f26 100644 --- a/src/kcat_tuning/evotune/loadEvotuneData.m +++ b/src/kcat_tuning/evotune/loadEvotuneData.m @@ -1,10 +1,11 @@ -function evoData = loadEvotuneData(varargin) +function experimentData = loadEvotuneData(varargin) % loadEvotuneData Load experimental data for evotune (CMA-ES) kcat tuning. % % Reads the three experimental data files used by cmaesKcatTuning from the -% data subfolder of the model directory defined by the model adapter. Each -% file is optional: a missing file yields [] (fluxData/maxGrate) or an -% empty cell array (zeroFlux) rather than an error. +% data subfolder of the model directory defined by the model adapter, with +% RAVEN's loadExperimentData. Each file is optional: a missing file yields [] +% (fluxData, maxGrowth) or an empty cell array (zeroFlux) rather than an +% error. % % Name-Value Arguments % -------------------- @@ -13,24 +14,24 @@ % % Returns % ------- -% evoData : struct +% experimentData : struct % structure with the loaded experimental data. % % Notes % ----- -% The evoData structure has the following fields: +% The experimentData structure has the following fields: % -% - fluxData : flux data loaded from evotuneFluxData.tsv, with the biomass -% reaction id stored in fluxData.biomass. [] if that file isn't present. -% - maxGrate : maximum growth-rate data loaded from evotuneMaxGrowth.tsv, -% with the biomass reaction id stored in maxGrate.biomass. [] if that -% file isn't present. +% - fluxData : flux data loaded from fluxData.tsv. Every condition in the +% file is used by the tuning; a weight of 0 in the column weight leaves a +% condition out. [] if the file isn't present. +% - maxGrowth : maximum growth-rate data loaded from maxGrowth.tsv. [] if +% the file isn't present. % - zeroFlux : reaction IDs assumed to carry zero flux, loaded from -% evotuneZeroExch.tsv. Empty cell array if that file isn't present. +% zeroFlux.tsv. Empty cell array if the file isn't present. % % See also % -------- -% cmaesKcatTuning, evotuneScore +% cmaesKcatTuning, evotuneScore, loadExperimentData p = parseRAVENargs(varargin, { ... 'modelAdapter', []}); modelAdapter = p.modelAdapter; @@ -42,32 +43,5 @@ end end -basePath = modelAdapter.params.path; - -evoData.fluxData = []; -fluxFile = fullfile(basePath,'data','evotuneFluxData.tsv'); -if isfile(fluxFile) - evoData.fluxData = loadFluxData(fluxFile, modelAdapter); -end -if ~isempty(evoData.fluxData) - evoData.fluxData.biomass = modelAdapter.params.bioRxn; -end - -evoData.maxGrate = []; -maxGrateFile = fullfile(basePath,'data','evotuneMaxGrowth.tsv'); -if isfile(maxGrateFile) - evoData.maxGrate = loadFluxData(maxGrateFile, modelAdapter); -end -if ~isempty(evoData.maxGrate) - evoData.maxGrate.biomass = modelAdapter.params.bioRxn; -end - -% All three files are optional: a missing file gives [] (fluxData, maxGrate) -% or {} (zeroFlux). -zeroExchFile = fullfile(basePath,'data','evotuneZeroExch.tsv'); -if isfile(zeroExchFile) - evoData.zeroFlux = table2cell(readtable(zeroExchFile, 'Delimiter', '\t', 'FileType','delimitedtext')); -else - evoData.zeroFlux = {}; -end +experimentData = loadExperimentData(fullfile(modelAdapter.params.path,'data')); end diff --git a/src/kcat_tuning/evotune/parsimonyFrontier.m b/src/kcat_tuning/evotune/parsimonyFrontier.m index fdd9932e..485703e2 100644 --- a/src/kcat_tuning/evotune/parsimonyFrontier.m +++ b/src/kcat_tuning/evotune/parsimonyFrontier.m @@ -14,9 +14,9 @@ % ---------- % scoreFcn : function_handle % scores one kcat vector; called once per threshold. Build it against -% a specific ecModel/evoData, e.g. +% a specific ecModel/experimentData, e.g. % scoreFcn = @(vec) evotuneScoreVector(ecModel,tunableIdx, ... -% ecRxnIdsTunable,evoData,modelAdapter,params,vec); +% ecRxnIdsTunable,experimentData,modelAdapter,params,vec); % where evotuneScoreVector applies vec to ecModel.ec.kcat(tunableIdx), % calls applyKcatConstraints, and returns evotuneScore's plain rmse. % kcat, kcat0, sigma0log : double diff --git a/src/kcat_tuning/evotune/screenKcatLeverage.m b/src/kcat_tuning/evotune/screenKcatLeverage.m index 95592b86..08800804 100644 --- a/src/kcat_tuning/evotune/screenKcatLeverage.m +++ b/src/kcat_tuning/evotune/screenKcatLeverage.m @@ -23,9 +23,9 @@ % -------------------- % modelAdapter : ModelAdapter % a loaded model adapter (default: the current default model adapter). -% evoData : struct +% experimentData : struct % as returned by loadEvotuneData (default: loaded from modelAdapter). -% Without fluxData and maxGrate, the error evotune:noData is raised. +% Without fluxData and maxGrowth, the error evotune:noData is raised. % tunableMask : logical % restrict which kcats are screened (default: every ec.kcat > 0). % fold : double @@ -49,12 +49,12 @@ % selectTunableMask, cmaesKcatTuning p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... - 'evoData', []; ... + 'experimentData', []; ... 'tunableMask', []; ... 'fold', 2.0; ... 'nProc', 1}); modelAdapter = p.modelAdapter; -evoData = p.evoData; +experimentData = p.experimentData; tunableMask = p.tunableMask; fold = p.fold; nProc = max(1, round(p.nProc)); @@ -65,12 +65,12 @@ error('Either send in a modelAdapter or set the default ecModel adapter in the ModelAdapterManager.') end end -if isempty(evoData) - evoData = loadEvotuneData('modelAdapter', modelAdapter); +if isempty(experimentData) + experimentData = loadEvotuneData('modelAdapter', modelAdapter); end -if isempty(evoData.fluxData) && isempty(evoData.maxGrate) - error('evotune:noData', ['evoData has neither fluxData nor maxGrate (from ' ... - 'data/evotuneFluxData.tsv and data/evotuneMaxGrowth.tsv); there is nothing to tune against.']) +if isempty(experimentData.fluxData) && isempty(experimentData.maxGrowth) + error('evotune:noData', ['experimentData has neither fluxData nor maxGrowth (from ' ... + 'data/fluxData.tsv and data/maxGrowth.tsv); there is nothing to tune against.']) end params = modelAdapter.params.evotune; @@ -96,18 +96,18 @@ nVec = numel(vectors); rmses = zeros(nVec,1); +% The first probe vector also gives the carbon weights (evotuneScore's 4th +% output), which are passed to every other call so the metabolite-formula +% scan behind them runs once for the whole screen rather than once per +% probe vector. +[~,rmses(1),~,carbonWeights] = scoreVector(ecModel,tunableIdx,ecRxnIdsTunable,experimentData,modelAdapter,params,vectors{1},[]); if nProc == 1 - % Serial path: pass the excarbon-augmented model from one call to - % the next, so the metabolite-formula scan behind it (evotuneScore's - % 4th output) runs once for the whole screen rather than once per - % probe vector. - baseModel = ecModel; - for v = 1:nVec - [~,rmses(v),~,baseModel] = scoreVector(baseModel,tunableIdx,ecRxnIdsTunable,evoData,modelAdapter,params,vectors{v}); + for v = 2:nVec + [~,rmses(v)] = scoreVector(ecModel,tunableIdx,ecRxnIdsTunable,experimentData,modelAdapter,params,vectors{v},carbonWeights); end else - parfor v = 1:nVec - [~,rmses(v)] = scoreVector(ecModel,tunableIdx,ecRxnIdsTunable,evoData,modelAdapter,params,vectors{v}); + parfor v = 2:nVec + [~,rmses(v)] = scoreVector(ecModel,tunableIdx,ecRxnIdsTunable,experimentData,modelAdapter,params,vectors{v},carbonWeights); end end @@ -149,8 +149,9 @@ 'VariableNames',{'rxnId','nIsozymes','sourceGroup','kcat0','leverage','sigma0log','rank','cumLeverageShare','positions'}); end -function [obj,rmse,rmseDetail,ecModel] = scoreVector(ecModel,tunableIdx,ecRxnIdsTunable,evoData,modelAdapter,params,kcatVec) +function [obj,rmse,rmseDetail,carbonWeights] = scoreVector(ecModel,tunableIdx,ecRxnIdsTunable,experimentData,modelAdapter,params,kcatVec,carbonWeights) ecModel.ec.kcat(tunableIdx) = kcatVec; ecModel = applyKcatConstraints(ecModel,'updateRxns',ecRxnIdsTunable); -[obj,rmse,rmseDetail,ecModel] = evotuneScore(ecModel,evoData,'modelAdapter',modelAdapter,'maxGrowthWeight',params.maxGrowthWeight); +[obj,rmse,rmseDetail,carbonWeights] = evotuneScore(ecModel,experimentData,'modelAdapter',modelAdapter,'maxGrowthWeight',params.maxGrowthWeight, ... + 'carbonWeights',carbonWeights); end diff --git a/src/kcat_tuning/evotune/tunePriorPenaltyWeight.m b/src/kcat_tuning/evotune/tunePriorPenaltyWeight.m index ee4c49f1..69df1ac2 100644 --- a/src/kcat_tuning/evotune/tunePriorPenaltyWeight.m +++ b/src/kcat_tuning/evotune/tunePriorPenaltyWeight.m @@ -26,7 +26,7 @@ % % Name-Value Arguments % -------------------- -% modelAdapter, evoData, tunableMask, screen, targetImpactShare, popsize, +% modelAdapter, experimentData, tunableMask, screen, targetImpactShare, popsize, % nProc, verbose % as in cmaesKcatTuning. % candidates : double @@ -56,7 +56,7 @@ % cmaesKcatTuning, screenKcatLeverage, selectTunableMask, nChanged p = parseRAVENargs(varargin, { ... 'modelAdapter', []; ... - 'evoData', []; ... + 'experimentData', []; ... 'tunableMask', []; ... 'screen', []; ... 'targetImpactShare', 0.9; ... @@ -67,7 +67,7 @@ 'seeds', [0, 1]; ... 'foldThreshold', 2.0}); modelAdapter = p.modelAdapter; -evoData = p.evoData; +experimentData = p.experimentData; tunableMask = p.tunableMask; screen = p.screen; targetImpactShare = p.targetImpactShare; @@ -84,13 +84,13 @@ error('Either send in a modelAdapter or set the default ecModel adapter in the ModelAdapterManager.') end end -if isempty(evoData) - evoData = loadEvotuneData('modelAdapter', modelAdapter); +if isempty(experimentData) + experimentData = loadEvotuneData('modelAdapter', modelAdapter); end if isempty(tunableMask) if isempty(screen) screen = screenKcatLeverage(ecModel, 'modelAdapter',modelAdapter, ... - 'evoData',evoData, 'nProc',nProc); + 'experimentData',experimentData, 'nProc',nProc); end tunableMask = selectTunableMask(ecModel, screen, 'targetImpactShare',targetImpactShare); end @@ -121,7 +121,7 @@ adapterCopy = modelAdapter; adapterCopy.params.evotune = lamParams; [~,res] = cmaesKcatTuning(ecModel, 'modelAdapter',adapterCopy, ... - 'evoData',evoData, 'tunableMask',tunableMask, ... + 'experimentData',experimentData, 'tunableMask',tunableMask, ... 'popsize',popsize, 'nProc',nProc, 'seed',seedVal, 'verbose',verbose); vectors{s} = res.newKcat; if ~isempty(res.rmseTrace) diff --git a/src/kcat_tuning/fillCarbonNum.m b/src/kcat_tuning/fillCarbonNum.m deleted file mode 100644 index 1e177bbe..00000000 --- a/src/kcat_tuning/fillCarbonNum.m +++ /dev/null @@ -1,327 +0,0 @@ -function model = fillCarbonNum(model) -% fillCarbonNum Add a carbon number field to the model. -% -% Adds a model.excarbon field with the carbon number of each exchange -% reaction, used to weight the RMSE during evotune kcat tuning. An exchange -% whose metabolite has no usable formula gets the carbon number 1. A model -% without a metFormulas field, or with only empty formulas, gives every -% exchange the carbon number 1 and a warning (GECKO:fillCarbonNum:noFormulas). -% -% Parameters -% ---------- -% model : struct -% ecModel that was generated by makeEcModel, or loaded from an earlier -% run. Not compatible with ecModels generated by earlier GECKO versions -% (pre 3.0). -% -% Returns -% ------- -% model : struct -% ecModel with an added model.excarbon field. - -% get carbonnum for each exchange rxn to further calculation of error -if ~isfield(model,'metFormulas') || all(cellfun(@isempty,model.metFormulas)) - warning('GECKO:fillCarbonNum:noFormulas', ... - 'The model has no metabolite formulas; every exchange reaction gets the carbon number 1.') - model.metFormulas = repmat({''},numel(model.mets),1); -end -model.excarbon = zeros(length(model.rxns),1); -[EXrxn, EXrxnIdx] = getExchangeRxns(model); -[CarbonNum,EXfors] = getcarbonnum(model,EXrxn); -model.excarbon(EXrxnIdx) = CarbonNum; -model.excarbon(strcmp(model.rxnNames,'growth')) = 41; - -function [CarbonNum,EXfors,changedmets] = getcarbonnum(model,exrxn) -% getcarbonnum Get carbon numbers matching the model. -% -% Parameters -% ---------- -% model : struct -% ecModel that was generated by makeEcModel, or loaded from an earlier -% run. Not compatible with ecModels generated by earlier GECKO versions -% (pre 3.0). -% exrxn : cell -% exchange reactions to use for counting. -% -% Returns -% ------- -% CarbonNum : double -% carbon number for each exchange reaction. -% EXfors : cell -% formulae of the exchanged metabolites. -% changedmets : cell -% exchange reactions whose carbon number could not be determined. -% - -% exrxn must be an exchange rxn and only contain one mets in the equation -[~,idx] = ismember(exrxn,model.rxns); -EXmets = model.S(:,idx); -% find all exchange mets -EXmetsIdx = zeros(length(exrxn),1); -for k = 1:length(EXmets(1,:)) - EXmetsIdx(k) = find(EXmets(:,k)); % -end -EXfors = model.metFormulas(EXmetsIdx); -[Ematrix, elements] = getElementalComposition(EXfors,{'C'}); -Ematrix = Ematrix(:,1); -changedmets = exrxn(isnan(Ematrix)); -Ematrix(isnan(Ematrix)) = 1; -CarbonNum = Ematrix; - -function [Ematrix, elements] = getElementalComposition(formulae, elements, chargeInFormula) -% Get the complete elemental composition matrix. It supports formulae with -% generic elements, parentheses and decimal places -% -% USAGE: -% [Ematrix, elements] = getElementalComposition(formulae, elements, chargeInFormula) -% -% INPUT: -% formulae: cell array of strings of chemical formulae. Can contain any generic elements starting -% with a capital letter followed by lowercase letters or '_', followed by a non-negative number. -% Also support '()', '[]', '{}'. E.g. {'H2O'; '[H2O]2(CuSO4)Generic_element0.5'} -% OPTIONAL INPUTS: -% elements: elements from previous call to preserve the order (default {}) -% chargeInFormula: true to accept formulae containing the generic element 'Charge' representing the charges, -% followed by a real number, e.g., 'HCharge1', 'SO4Charge-2' (default false). -% -% OUTPUTS: -% Ematrix: elemental composition matrix (#formulae x #elements) -% elements: cell array of elements corresponding to the columns of Ematrix -% -% E.g., [Ematrix, elements] = getElementalComposition({'H2O'; '[H2O]2(CuSO4)Generic_element0.5'}) would return: -% elements = {'H', 'O', 'Cu', 'S', 'Generic_element'} -% Ematrix = [ 2, 1, 0, 0, 0; -% 4, 6, 1, 1, 0.5] -% -% Siu Hung Joshua Chan May 2017 - -if nargin < 3 || isempty(chargeInFormula) - chargeInFormula = false; -else - chargeInFormula = logical(chargeInFormula); -end -% for recalling the original formula at the top level if there are parentheses -% in the formula leading to iterative calling -persistent formTopLv -persistent formCurLv -% for storing error message during iterative calling -persistent errMsg -persistent errMsgInThisLoop -persistent topLvJ -persistent selfCall -if isempty(selfCall) - selfCall = 0; -end -if ~selfCall - [formTopLv, errMsg] = deal(''); -end - -if ~isstruct(formulae) - if ~iscell(formulae) - % make sure it is a cell array of strings - formulae = {formulae}; - end -else - % also accept COBRA model as input - if isfield(formulae, 'metFormulas') - formulae = formulae.metFormulas; - else - error('The 1st input ''formulae'' should be a cell array of strings of formulae or a COBRA model with *.metFormulas') - end -end - -if nargin < 2 || isempty(elements) - elements = {}; -elseif numel(unique(elements)) < numel(elements) - error('Repeated elements in the input ''elements'' array.') -else - elements = elements(:)'; % make sure it is a row vector -end - -% the Ematrix -Ematrix = zeros(numel(formulae), numel(elements)); -% replace all brackets and braces by parentheses -formulae = regexprep(formulae, '[\[\{]', '\('); -formulae = regexprep(formulae, '[\]\}]', '\)'); -digit = floor(log10(numel(formulae))) + 1; -for j = 1:numel(formulae) - if ~selfCall - % reset top level information for each formula - [formTopLv, topLvJ, errMsgInThisLoop] = deal(formulae{j}, j, false); - end - formulae{j} = strtrim(formulae{j}); - if ~isempty(formulae{j}) - % get all outer parentheses - parenthesis = []; - stP = []; - stPpos = []; - lv = 0; - k = 1; - while k <= length(formulae{j}) - if strcmp(formulae{j}(k),'(') - if lv == 0 - pStart = k; - end - lv = lv + 1; - elseif strcmp(formulae{j}(k),')') - if lv == 1 - parenthesis = [parenthesis; [pStart k]]; - % right parenthesis, get the following stoichiometry if any - stPre = regexp(formulae{j}(k+1:end), '^(\+|\-)?\d*\.?\d*', 'match'); - if isempty(stPre) - stP = [stP; 1]; - stPpos = [stPpos; k k]; - else - stCheck = str2double(stPre{1}); - s = ''; - if isnan(stCheck) - s = 'Invalid'; - elseif stCheck < 0 - s = 'Negative'; - end - if ~isempty(s) % error if not convertible to number or negative - f = [formulae{j}(parenthesis(end,1):parenthesis(end,2)), stPre{1}]; - addErrorMessage(sprintf(' %s stoichiometry in ''%s''\n', s, f)) - end - % stoichiometry - stP = [stP; stCheck]; - % position of the stoichiometry in the text - stPpos = [stPpos; k + 1, k + length(stPre{1})]; - k = k + length(stPre{1}); - end - end - lv = lv - 1; - end - k = k + 1; - end - if isempty(parenthesis) - % No parenthesis, parse the formula - re = regexp(formulae{j}, '([A-Z][a-z_]*)((?:\+|\-)?\d*\.?\d*)', 'tokens'); - s = strjoin(cellfun(@(x) strjoin(x,''), re, 'UniformOutput', false), ''); - errorFlag = 0; - if ~strcmp(s, formulae{j}) % check if the entire formula is retrieved - errorFlag = 1; - else - % check the stoichiometry - eleCheck = cellfun(@(x) x{1}, re, 'UniformOutput', false); - stCheck = cellfun(@(x) str2double(x{2}), re); - stCheck(cellfun(@(x) isempty(x{2}), re)) = 1; % empty value => stoichiometry = 1 - if any(isnan(stCheck)) - f = strjoin(cellfun(@(x) strjoin(x, ''), re(isnan(stCheck)), 'UniformOutput', false), ''', '''); - [errorFlag, errMsgKey] = deal(2, 'Invalid'); - else - % if chargeInFormula is true, detect negative stoich for non-charge elements, else any negative stoich - negSt = (~chargeInFormula | ~strcmp(eleCheck, 'Charge')) & stCheck < 0; - if any(negSt) - f = strjoin(cellfun(@(x) strjoin(x, ''), re(negSt), 'UniformOutput', false), ''', '''); - [errorFlag, errMsgKey] = deal(2, 'Negative'); - end - end - end - if errorFlag > 0 - if selfCall <= 1 - s2 = sprintf('from the part ''%s''', formulae{j}); - elseif selfCall == 2 - s2 = sprintf('from the part ''(%s)''', formCurLv); - elseif selfCall == 3 - s2 = sprintf('from the part ''(%s)''', formulae{j}); - end - if errorFlag == 1 - s2 = sprintf([' Only ''%s'' can be recognized %s.\n'... - ' Each element should start with a capital letter followed by lower case letters'... - ' or ''_'' with indefinite length and followed by a number.\n'], s, s2); - elseif errorFlag == 2 - s2 = sprintf(' %s stoichiometry in ''%s'' %s\n', errMsgKey, f, s2); - end - addErrorMessage(s2); - end - elementJ = repmat({''}, 1, numel(re)); - nEj = 0; - stoichJ = zeros(numel(re),1); - for k = 1:numel(re) - [ynK,idK] = ismember(re{k}(1), elementJ(1:nEj)); - if ynK - k2 = idK; - else - nEj = nEj + 1; - elementJ{nEj} = re{k}{1}; - k2 = nEj; - end - if isempty(re{k}{2}) - stoichJ(k2) = stoichJ(k2) + 1; - else - stoichJ(k2) = stoichJ(k2) + str2double(re{k}{2}); - end - end - elementJ = elementJ(1:nEj); - stoichJ = stoichJ(1:nEj); - [ynE, idE] = ismember(elementJ, elements); % map to existing elements - if any(~ynE) - idE(~ynE) = (numel(elements) + 1):(numel(elements) + sum(~ynE)); - Ematrix(:, (numel(elements) + 1):(numel(elements) + sum(~ynE))) = 0; - elements = [elements, elementJ(~ynE)]; - end - Ematrix(j, idE) = stoichJ; - else - % parentheses found. Iteratively get the formula inside parentheses - rest = true(length(formulae{j}),1); - for k = 1:size(parenthesis,1) - selfCallCur = selfCall; - selfCall = 3; - [EmatrixK, elements] = getElementalComposition(formulae{j}(... - (parenthesis(k,1)+1):(parenthesis(k,2)-1)), elements, chargeInFormula); - if numel(elements) > size(Ematrix, 2) - Ematrix(:, (size(Ematrix, 2) + 1):numel(elements)) = 0; - end - selfCall = selfCallCur; - Ematrix(j, 1:numel(elements)) = Ematrix(j, 1:numel(elements)) + EmatrixK * stP(k); - rest(parenthesis(k,1):stPpos(k,2)) = false; - end - if any(rest) - formCurLv = formulae{j}; - selfCallCur = selfCall; - selfCall = 1 + (selfCall >= 1); - while any(rest) - % get the consecutive part of the formula that is not in parentheses - [pStart, pEnd] = deal(find(rest, 1), find(~rest)); - pEnd = min(pEnd(pEnd > pStart)); - if isempty(pEnd) - pEnd = numel(rest); - else - pEnd = pEnd - 1; - end - [EmatrixK, elements] = getElementalComposition(formulae{j}(pStart:pEnd), elements, chargeInFormula); - if numel(elements) > size(Ematrix, 2) - Ematrix(:, (size(Ematrix, 2) + 1):numel(elements)) = 0; - end - Ematrix(j, 1:numel(elements)) = Ematrix(j, 1:numel(elements)) + EmatrixK; - rest(pStart:pEnd) = false; - end - selfCall = selfCallCur; - end - end - else - Ematrix(j,:) = NaN; - end -end -Ematrix = Ematrix(:, 1:numel(elements)); -if ~selfCall - selfCall = []; - if ~isempty(errMsg) - error(['%s\n', errMsg], 'Invalid formula input:') - end -end - -% nested function for adding error messages - function addErrorMessage(s) - if ~errMsgInThisLoop - errMsg = [errMsg, sprintf(['#%0' num2str(digit) 'd: %s\n'], topLvJ, formTopLv)]; - errMsgInThisLoop = true; - end - errMsg = [errMsg, s]; - end -end - -end - -end diff --git a/src/limit_proteins/constrainFluxData.m b/src/limit_proteins/constrainFluxData.m index 460e3471..18a13184 100644 --- a/src/limit_proteins/constrainFluxData.m +++ b/src/limit_proteins/constrainFluxData.m @@ -2,7 +2,9 @@ % constrainFluxData Constrain ecModel fluxes to provided flux data. % % Constrains fluxes to the data that is provided in the fluxData structure, -% which itself is read by loadFluxData from data/fluxData.tsv. +% which itself is read by loadFluxData from data/fluxData.tsv. The bounds are +% set by RAVEN's applyFluxDataConstraints; this function takes the biomass +% reaction and the preferred carbon source from the model adapter. % % Parameters % ---------- @@ -20,6 +22,8 @@ % - exchFluxes : exchange fluxes (mmol/gDCWh). % - exchMets : exchanged metabolites, matching exchFluxes. % - exchRxnIDs : exchange reaction IDs, matching exchMets. +% +% Default: the data read from data/fluxData.tsv of the model adapter. % condition : double or char % either index number or name of the sample condition in fluxData.conds % (default 1). @@ -51,8 +55,10 @@ % Notes % ----- % An exchange reaction in fluxData.exchRxnIDs that is not in the model -% raises constrainFluxData:missingReaction, and a condition that is not in -% fluxData.conds raises constrainFluxData:unknownCondition. +% raises applyFluxDataConstraints:missingReaction, and a condition that is not +% in fluxData.conds raises applyFluxDataConstraints:unknownCondition. The +% preferred carbon source of the adapter (c_source) is blocked before the +% data is applied. % % If a provided constraint is either -1000 or 1000, then the function will % update the reaction lower and upper bound to either allow uptake or @@ -67,7 +73,7 @@ % % See also % -------- -% loadFluxData +% loadFluxData, applyFluxDataConstraints p = parseRAVENargs(varargin, { ... 'fluxData', []; ... @@ -76,9 +82,6 @@ 'looseStrictFlux', []; ... 'modelAdapter', []}); fluxData = p.fluxData; -condition = p.condition; -maxMinGrowth = p.maxMinGrowth; -looseStrictFlux = p.looseStrictFlux; modelAdapter = p.modelAdapter; if isempty(modelAdapter) @@ -89,83 +92,19 @@ end params = modelAdapter.getParameters(); -if isempty(looseStrictFlux) - looseStrictFlux = 'loose'; -end - -if isempty(maxMinGrowth) - maxMinGrowth = 'max'; -end - if isempty(fluxData) - fluxData = loadFluxData(fullfile(params.path,'data','fluxData.tsv'),modelAdapter); + fluxData = loadFluxData(fullfile(params.path,'data','fluxData.tsv')); end -if ~(ischar(maxMinGrowth) && any(strcmp(maxMinGrowth, {'max','min'}))) - error('constrainFluxData:invalidMaxMinGrowth', ... - 'maxMinGrowth must be ''max'' or ''min''.') +options = {'cSource', params.c_source}; +if ~isempty(p.condition) + options = [options, {'condition', p.condition}]; end - -missingRxns = fluxData.exchRxnIDs(~ismember(fluxData.exchRxnIDs, model.rxns)); -if ~isempty(missingRxns) - error('constrainFluxData:missingReaction', ... - 'The exchange reactions %s of fluxData are not in the model.', strjoin(missingRxns, ', ')) -end - -if isempty(condition) - condition = 1; -elseif ~isnumeric(condition) - idx = find(strcmp(fluxData.conds,condition)); - if isempty(idx) - error('constrainFluxData:unknownCondition', ... - 'Condition %s cannot be found in fluxData.', condition) - else - condition = idx; - end +if ~isempty(p.maxMinGrowth) + options = [options, {'maxMinGrowth', p.maxMinGrowth}]; end - -% Select the exchange fluxes of specified condition -fluxData.exchFluxes = fluxData.exchFluxes(condition,:); - -%Set original c-source to zero -model = setParam(model,'eq',params.c_source,0); -%Set growth -switch maxMinGrowth - case 'max' - model = setParam(model,'lb',params.bioRxn,0); - model = setParam(model,'ub',params.bioRxn,fluxData.grRate(condition)); - case 'min' - model = setParam(model,'lb',params.bioRxn,fluxData.grRate(condition)); - model = setParam(model,'ub',params.bioRxn,1000); -end - -naFlux = isnan(fluxData.exchFluxes); -negFlux = lt(fluxData.exchFluxes,0); % less than 0 -ub = fluxData.exchFluxes(~(negFlux | naFlux)); -posFlux = fluxData.exchRxnIDs(~(negFlux | naFlux)); -lb = fluxData.exchFluxes(negFlux & ~naFlux); -negFlux = fluxData.exchRxnIDs(negFlux & ~naFlux); - -switch looseStrictFlux - case 'loose' - model = setParam(model,'lb',negFlux,lb); - model = setParam(model,'ub',negFlux,0); - model = setParam(model,'lb',posFlux,0); - model = setParam(model,'ub',posFlux,ub); - otherwise - model = setParam(model,'var',negFlux,lb,looseStrictFlux); - model = setParam(model,'var',posFlux,ub,looseStrictFlux); -end -extremeFlux = find(abs(fluxData.exchFluxes)==1000); -if any(extremeFlux) - exchFlux = fluxData.exchFluxes(extremeFlux); - if any(exchFlux==-1000) - model = setParam(model,'lb',fluxData.exchRxnIDs(extremeFlux(exchFlux==-1000)),-1000); - model = setParam(model,'ub',fluxData.exchRxnIDs(extremeFlux(exchFlux==-1000)),0); - end - if any(exchFlux==1000) - model = setParam(model,'lb',fluxData.exchRxnIDs(extremeFlux(exchFlux==1000)),0); - model = setParam(model,'ub',fluxData.exchRxnIDs(extremeFlux(exchFlux==1000)),1000); - end +if ~isempty(p.looseStrictFlux) + options = [options, {'looseStrictFlux', p.looseStrictFlux}]; end +model = applyFluxDataConstraints(model, fluxData, params.bioRxn, options{:}); end diff --git a/src/utilities/upgradeGECKOproject.m b/src/utilities/upgradeGECKOproject.m index 1630c999..7df131ec 100644 --- a/src/utilities/upgradeGECKOproject.m +++ b/src/utilities/upgradeGECKOproject.m @@ -205,7 +205,7 @@ function keepOriginal(file) 'enzymeUsage', 'getEnzymeUsage'; ... 'selectKcatValue', 'assignKcatValues'; 'startGECKOproject', 'createGECKOproject'; ... 'updateGECKOdoc', 'buildGECKOdoc'; 'copyECtoGEM', 'applyECcodes'; ... - 'addCarbonNum', 'fillCarbonNum'; 'getcarbonnum', 'fillCarbonNum'; ... + 'addCarbonNum', 'computeCarbonWeights (RAVEN)'; 'getcarbonnum', 'computeCarbonWeights (RAVEN)'; ... 'bayesianSensitivityTuning', 'cmaesKcatTuning'; 'abc_max', 'cmaesKcatTuning'; ... 'getrSample', 'cmaesKcatTuning'; 'updateprior', 'cmaesKcatTuning'; ... 'findMaxValue', 'kcatsFromBrenda (or kcatsFromECTree)'; ... diff --git a/test/VerificationMatrix.tsv b/test/VerificationMatrix.tsv index ffb796f5..caa1de94 100644 --- a/test/VerificationMatrix.tsv +++ b/test/VerificationMatrix.tsv @@ -41,7 +41,6 @@ gather_kcats/kcatsFromHomologs.m tc0095, tc0096 gather_kcats/extractEnzymeSubstratePairs.m tc0096, tc0015 gather_kcats/convertCustomKcats.m tc0091, tc0092, tc0094, tc0725 change_model/upgradeEcModel.m tc0093, tc0094, tc0099 -io/loadFluxData.m tc0506 io/loadProtData.m tc0101, tc0102, tc0260, tc0298, tc0299, tc0313 change_model/addNewRxnsToEC.m tc0211, tc0232, tc0237, tc0500 change_model/mapRxnsToConv.m tc0213 @@ -70,9 +69,9 @@ download_databases/fetchBrendaAsset.m tc0085 download_databases/loadDatabases.m Called from getECfromDatabase - tested when that function is tested., tc0729 kcat_tuning/evotune/bestParsimonious.m tc0600 kcat_tuning/evotune/* Has only been tested by running it manually, not tested by test cases -kcat_tuning/fillCarbonNum.m tc0055 (deprecated-alias check only), tc0601, tc0602 +kcat_tuning/evotune/loadEvotuneData.m tc0601, tc0602 kcat_tuning/evotune/* resolveTunableContext by tc0312, tc0719; the evotune:noData check of cmaesKcatTuning and screenKcatLeverage by tc0709; the rest has only been tested by running it manually, not tested by test cases -kcat_tuning/fillCarbonNum.m tc0055 (deprecated-alias check only, not a real usage test) +kcat_tuning/evotune/evotuneScore.m not tested by a test case; the conditions, the distance and the carbon weights it calls are tested in RAVEN (tExperiments) kcat_tuning/findTopLimitations.m Not explicitly tested - The code has existed for a long time kcat_tuning/sensitivityTuning.m tcm0001, tc0251, tc0253, tc0254 kcat_tuning/sigmaFitter.m tc0020, tc0255, tc0710 diff --git a/test/unit_tests/geckoCoreFunctionTests.m b/test/unit_tests/geckoCoreFunctionTests.m index 9b92ea06..89ff9890 100644 --- a/test/unit_tests/geckoCoreFunctionTests.m +++ b/test/unit_tests/geckoCoreFunctionTests.m @@ -2383,15 +2383,8 @@ function testDeprecatedAliasesWarnAndForward_tc0055(testCase) resultNew = applyECcodes(ecCopy); verifyEqual(testCase, resultOld.eccodes, resultNew.eccodes) - % addCarbonNum -> fillCarbonNum - carbonModel = getGeckoTestModel(); - carbonModel.metFormulas = {'C6H12O6';'C2H6O';'C6H12O6';'C2H6O'}; - lastwarn('',''); - carbonOld = addCarbonNum(carbonModel); - [~,warnId] = lastwarn(); - verifyEqual(testCase, warnId, 'GECKO:deprecatedName') - carbonNew = fillCarbonNum(carbonModel); - verifyEqual(testCase, carbonOld.excarbon, carbonNew.excarbon) + % addCarbonNum was removed; the error names computeCarbonWeights + verifyError(testCase, @() addCarbonNum(getGeckoTestModel()), 'GECKO:removedFunction') end function testGetStandardKcatSubsystemValueIsMedianNotMean_tc0056(testCase) @@ -2764,11 +2757,11 @@ function testEvotuneWithoutDataErrors_tc0709(testCase) % against: screenKcatLeverage and cmaesKcatTuning raise evotune:noData. [ecModel, adapter, project] = buildScratchProject(); cleanupDir = onCleanup(@() rmdir(project, 's')); - evoData = struct('fluxData', [], 'maxGrate', [], 'zeroFlux', {{}}); + experimentData = struct('fluxData', [], 'maxGrowth', [], 'zeroFlux', {{}}); verifyError(testCase, @() screenKcatLeverage(ecModel, 'modelAdapter', adapter, ... - 'evoData', evoData), 'evotune:noData') + 'experimentData', experimentData), 'evotune:noData') verifyError(testCase, @() cmaesKcatTuning(ecModel, 'modelAdapter', adapter, ... - 'evoData', evoData), 'evotune:noData') + 'experimentData', experimentData), 'evotune:noData') % The test project has no evotune files, so the loaded data is empty too verifyError(testCase, @() screenKcatLeverage(ecModel, 'modelAdapter', adapter), 'evotune:noData') end @@ -3193,7 +3186,7 @@ function testRemovedGECKO3FunctionsNameTheirReplacement_tc0731(testCase) % The stubs of removed GECKO 3 functions raise an error that names the % GECKO 4 replacement. cases = {@abc_max, 'cmaesKcatTuning'; @getrSample, 'cmaesKcatTuning'; ... - @updateprior, 'cmaesKcatTuning'; @getcarbonnum, 'fillCarbonNum'; ... + @updateprior, 'cmaesKcatTuning'; @getcarbonnum, 'computeCarbonWeights'; ... @findMaxValue, 'kcatsFromBrenda'}; for k = 1:size(cases, 1) try diff --git a/test/unit_tests/parityRound5G3Tests.m b/test/unit_tests/parityRound5G3Tests.m index 5be6ba8d..f6259a4c 100644 --- a/test/unit_tests/parityRound5G3Tests.m +++ b/test/unit_tests/parityRound5G3Tests.m @@ -89,41 +89,13 @@ function testConstrainFluxDataInvalidArguments_tc0505(testCase) constrained = constrainFluxData(ecModel, 'fluxData', fluxData, 'modelAdapter', adapter); verifyEqual(testCase, constrained.ub(strcmp(constrained.rxns, 'S1')), 1) verifyError(testCase, @() constrainFluxData(ecModel, 'fluxData', fluxData, ... - 'maxMinGrowth', 'median', 'modelAdapter', adapter), 'constrainFluxData:invalidMaxMinGrowth') + 'maxMinGrowth', 'median', 'modelAdapter', adapter), 'applyFluxDataConstraints:invalidMaxMinGrowth') fluxData.exchRxnIDs = {'NOT_A_RXN'}; verifyError(testCase, @() constrainFluxData(ecModel, 'fluxData', fluxData, ... - 'modelAdapter', adapter), 'constrainFluxData:missingReaction') + 'modelAdapter', adapter), 'applyFluxDataConstraints:missingReaction') fluxData.exchRxnIDs = {'S1'}; verifyError(testCase, @() constrainFluxData(ecModel, 'fluxData', fluxData, ... - 'condition', 'nope', 'modelAdapter', adapter), 'constrainFluxData:unknownCondition') -end - -function testLoadFluxDataMalformedFiles_tc0506(testCase) - adapter = getAdapter(); - folder = tempname; - mkdir(folder) - cleanup = onCleanup(@() rmdir(folder, 's')); - tab = char(9); - - verifyError(testCase, @() loadFluxData(fullfile(folder, 'missing.tsv'), adapter), ... - 'loadFluxData:fileNotFound') - - fixedOnly = writeLines(fullfile(folder, 'fixed.tsv'), ... - {['cond' tab 'Ptot' tab 'grRate'], ['c1' tab '0.5' tab '0.4']}); - verifyError(testCase, @() loadFluxData(fixedOnly, adapter), 'loadFluxData:noExchangeColumns') - - ragged = writeLines(fullfile(folder, 'ragged.tsv'), ... - {['cond' tab 'Ptot' tab 'grRate' tab 'Glc (EX_a)' tab 'O2 (EX_b)'], ... - ['c1' tab '0.5' tab '0.4' tab '-1' tab '2'], ... - ['c2' tab '0.5' tab '0.4' tab '-1']}); - verifyError(testCase, @() loadFluxData(ragged, adapter), 'loadFluxData:raggedRow') - - nested = writeLines(fullfile(folder, 'nested.tsv'), ... - {['cond' tab 'Ptot' tab 'grRate' tab 'X (EX_a(e))' tab 'O2 (EX_b)'], ... - ['c1' tab '0.5' tab '0.4' tab '-1' tab '2']}); - fluxData = loadFluxData(nested, adapter); - verifyEqual(testCase, fluxData.exchMets, {'X', 'O2'}) - verifyEqual(testCase, fluxData.exchRxnIDs, {'EX_a(e)', 'EX_b'}) + 'condition', 'nope', 'modelAdapter', adapter), 'applyFluxDataConstraints:unknownCondition') end function testGetEnzymeUsageLeavesOutMissingUsageReaction_tc0507(testCase) @@ -225,9 +197,3 @@ function testApplyKcatConstraintsLightUpdateIds_tc0512(testCase) geckoPath = findGECKOroot; adapter = ModelAdapterManager.getAdapter(fullfile(geckoPath,'test','unit_tests','ecTestGEM','TestGEMAdapter.m')); end - -function file = writeLines(file, lines) -fid = fopen(file, 'w'); -fprintf(fid, '%s\n', lines{:}); -fclose(fid); -end diff --git a/test/unit_tests/parityRound5G4Tests.m b/test/unit_tests/parityRound5G4Tests.m index 61c5aecd..1086db6f 100644 --- a/test/unit_tests/parityRound5G4Tests.m +++ b/test/unit_tests/parityRound5G4Tests.m @@ -13,24 +13,34 @@ function testBestParsimoniousSkipsNaNRmse_tc0600(testCase) verifyEqual(testCase, point.nChanged, 3000) end -function testFillCarbonNumWithoutFormulasWarnsAndWeightsOne_tc0601(testCase) - model = exchangeToyModel(); - verifyWarning(testCase, @() fillCarbonNum(model), 'GECKO:fillCarbonNum:noFormulas') - model.metFormulas = {'';''}; - verifyWarning(testCase, @() fillCarbonNum(model), 'GECKO:fillCarbonNum:noFormulas') - warning('off','GECKO:fillCarbonNum:noFormulas') - cleanup = onCleanup(@() warning('on','GECKO:fillCarbonNum:noFormulas')); - withoutField = rmfield(model,'metFormulas'); - out = fillCarbonNum(withoutField); - verifyEqual(testCase, out.excarbon(strcmp(out.rxns,'EX_a')), 1) - verifyEqual(testCase, out.excarbon(strcmp(out.rxns,'EX_b')), 1) +function testLoadEvotuneDataReadsTheExperimentFiles_tc0601(testCase) + adapter = ModelAdapterManager.getAdapter(fullfile(findGECKOroot,'test','unit_tests','ecTestGEM','TestGEMAdapter.m')); + project = tempname; + mkdir(fullfile(project, 'data')); + testCase.addTeardown(@() rmdir(project, 's')); + adapter.params.path = project; + tab = char(9); + writeLines(fullfile(project, 'data', 'fluxData.tsv'), ... + {['Condition' tab 'Ptot' tab 'grRate' tab 'glucose (EX_glc)' tab 'weight'], ... + ['c1' tab '0.5' tab '0.4' tab '-10' tab '0'], ... + ['c2' tab '0.5' tab '0.3' tab '-8' tab '1']}); + writeLines(fullfile(project, 'data', 'maxGrowth.tsv'), ... + {['Condition' tab 'Ptot' tab 'grRate' tab 'glucose (EX_glc)'], ... + ['glucose' tab 'NaN' tab '0.41' tab '-1000']}); + writeLines(fullfile(project, 'data', 'zeroFlux.tsv'), {'Rxns', 'EX_a', 'EX_b'}); + experimentData = loadEvotuneData('modelAdapter', adapter); + verifyEqual(testCase, experimentData.fluxData.conds, {'c1'; 'c2'}) + verifyEqual(testCase, experimentData.fluxData.weight, [0; 1]) + verifyEqual(testCase, experimentData.maxGrowth.exchRxnIDs, {'EX_glc'}) + verifyEqual(testCase, experimentData.zeroFlux, {'EX_a'; 'EX_b'}) end -function testFillCarbonNumUsesFormulaCarbons_tc0602(testCase) - model = exchangeToyModel(); - model.metFormulas = {'C6H12O6';'O2'}; - out = fillCarbonNum(model); - verifyEqual(testCase, out.excarbon(strcmp(out.rxns,'EX_a')), 6) +function testLoadEvotuneDataWithoutFilesIsEmpty_tc0602(testCase) + adapter = ModelAdapterManager.getAdapter(fullfile(findGECKOroot,'test','unit_tests','ecTestGEM','TestGEMAdapter.m')); + experimentData = loadEvotuneData('modelAdapter', adapter); + verifyEmpty(testCase, experimentData.fluxData) + verifyEmpty(testCase, experimentData.maxGrowth) + verifyEmpty(testCase, experimentData.zeroFlux) end function testSaveEcModelWritesDefaultDescriptionOnlyWhenNone_tc0603(testCase) @@ -75,3 +85,9 @@ function testSaveEcModelWritesDefaultDescriptionOnlyWhenNone_tc0603(testCase) model.compNames = {'cytosol'}; model.metComps = [1;1]; end + +function file = writeLines(file, lines) +fid = fopen(file, 'w'); +fprintf(fid, '%s\n', lines{:}); +fclose(fid); +end diff --git a/tutorials/full_ecModel/data/evotuneFluxData.tsv b/tutorials/full_ecModel/data/evotuneFluxData.tsv deleted file mode 100644 index 67aca488..00000000 --- a/tutorials/full_ecModel/data/evotuneFluxData.tsv +++ /dev/null @@ -1,34 +0,0 @@ -Condition Ptot grRate glucose (r_1714) fructose (r_1709) sucrose (r_2058) xylose (r_1718) CO2 (r_1672) oxygen (r_1992) pyruvate (r_2033) ethyl acetate (r_1765) glycerol (r_1808) acetate (r_1634) ethanol (r_1761) succinate (r_2056) evotuneRMSEweight source -glucose NaN 0.36 -13.33333333 0 0 0 NaN NaN 0 0 0.38 NaN 19.82608696 NaN 1 DLKcat -glucose NaN 0.39 -15.44444444 0 0 0 NaN NaN 0 0 0.37 NaN 22.4068323 NaN 1 DLKcat -glucose NaN 0.37 -19.55555556 0 0 0 NaN NaN 0 0 NaN NaN NaN NaN 1 DLKcat -glucose NaN 0.37 -14.8 0 0 0 NaN NaN 0 0 1.1 0.4 22 NaN 1 DLKcat -glucose NaN 0.38 -15.05555556 0 0 0 NaN NaN 0 0 NaN NaN NaN NaN 1 DLKcat -glucose NaN 0.35 -17.66666667 0 0 0 NaN NaN 0 0 0.34 NaN 23.51778656 NaN 1 DLKcat -glucose NaN 0.37 -11.2 0 0 0 19.58 NaN 0 0 0.12 0.43 15.74 NaN 1 DLKcat -glucose NaN 0.37 -13.66666667 0 0 0 NaN NaN 0 0 NaN NaN NaN NaN 1 DLKcat -glucose NaN 0.38 -17.55555556 0 0 0 NaN NaN 0 0 0.41 NaN 28.22463768 NaN 1 DLKcat -glucose NaN 0.32 -25.38888889 0 0 0 NaN NaN 0 0 0.43 NaN 42.73291925 NaN 1 DLKcat -fructose NaN 0.1 0 -6.888888889 0 0 NaN 0 0 0 0.41 NaN 11.14130435 NaN 1 DLKcat -fructose NaN 0.25 0 -12.44444444 0 0 NaN 0 0 0 0.41 NaN 20.256917 NaN 1 DLKcat -sucrose NaN 0.32 0 0 -6.076541046 0 NaN 0 0 0 NaN NaN 19.94202899 NaN 1 DLKcat -glucose NaN 0.025 -0.3 0 0 0 0.75198 -0.73102 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 -glucose NaN 0.05 -0.6 0 0 0 1.3697 -1.3251 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 -glucose NaN 0.1 -1.1 0 0 0 2.4007 -2.3167 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 -glucose NaN 0.15 -1.7 0 0 0 3.6379 -3.5066 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 -glucose NaN 0.2 -2.3 0 0 0 4.8751 -4.6966 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 -glucose NaN 0.25 -2.8 0 0 0 5.906 -5.6881 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 -glucose NaN 0.28 -3.4 0 0 0 7.0413 -6.6773 0 0 0 0.08 0.11 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 -glucose NaN 0.3 -4.5 0 0 0 8.5667 -6.012 0 0 0 0.41 2.3 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 -glucose NaN 0.35 -8.6 0 0 0 15.2091 -5.3809 0 0 0 0.62 9.5 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 -glucose NaN 0.4 -11.1 0 0 0 19.3207 -5.0171 0 0 0 0.6 13.9 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 -glucose NaN 0.4 -20.2 0 0 0 34.4 NaN 0 0 0 1 30 NaN 1 DLKcat -glucose NaN 0.41 -14.22222222 0 0 0 NaN NaN 0 0 0 NaN 25.625 NaN 1 DLKcat -glucose NaN 0.1 -6.44 0 0 0 NaN 0 0 0 1.08 0 9.26 NaN 1 doi:10.1186/1752-0509-2-60 -glucose NaN 0.2449 -10.173 0 0 0 19.109 NaN 0 0 1.2908 0.212 15.494 NaN 1 doi:10.1002/bit.25447 -glucose NaN 0.2287 -13.492 0 0 0 27.31 0 0 0 2.9849 0.328 20.636 NaN 1 doi:10.1002/bit.25447 -glucose NaN 0.1 -6.0184 0 0 0 9.8220288 0 0.0120368 0 1.0351648 0.0240736 8.9734344 0.0270828 1 doi:10.1099/00221287-143-1-203 -glucose NaN 0.2 -11.5927 0 0 0 18.9888426 0 0.0463708 0 2.1098714 0.0695562 17.2499376 0.05216715 1 doi:10.1099/00221287-143-1-203 -glucose NaN 0.3 -17.0704 0 0 0 27.3467808 0 0.1024224 0 3.243376 0.2048448 25.2983328 0.0512112 1 doi:10.1099/00221287-143-1-203 -glucose NaN 0.4 -22.5483 0 0 0 35.3106378 0 0.1803864 0 4.9155294 0.450966 33.6195153 0.0676449 1 doi:10.1099/00221287-143-1-203 -glucose 0.46 0.1 -1.0709285 0 0 0 2.627555 -2.446473 NaN NaN 0.0008575 0 0 NaN 1 doi:10.1016/j.cels.2017.03.003 diff --git a/tutorials/full_ecModel/data/evotuneMaxGrowth.tsv b/tutorials/full_ecModel/data/evotuneMaxGrowth.tsv deleted file mode 100644 index 8281de10..00000000 --- a/tutorials/full_ecModel/data/evotuneMaxGrowth.tsv +++ /dev/null @@ -1,9 +0,0 @@ -Condition Ptot grRate glucose (r_1714) fructose (r_1709) sucrose (r_2058) acetate (r_1634) ethanol (r_1761) galactose (r_1710) maltose (r_1931) mannose (r_1715) evotuneRMSEweight source -fructose NaN 0.338 NaN -1000 NaN NaN NaN NaN NaN NaN 1 DLKcat -glucose NaN 0.41 -1000 NaN NaN NaN NaN NaN NaN NaN 1 DLKcat -ethanol NaN 0.12 NaN NaN NaN NaN -1000 NaN NaN NaN 1 DLKcat -mannose NaN 0.33 NaN NaN NaN NaN NaN NaN NaN -1000 1 DLKcat -galactose NaN 0.28 NaN NaN NaN NaN NaN -1000 NaN NaN 1 DLKcat -acetate NaN 0.17 NaN NaN NaN -1000 NaN NaN NaN NaN 1 DLKcat -maltose NaN 0.4 NaN NaN NaN NaN NaN NaN -1000 NaN 1 DLKcat -sucrose NaN 0.39 NaN NaN -1000 NaN NaN NaN NaN NaN 1 DLKcat diff --git a/tutorials/full_ecModel/data/fluxData.tsv b/tutorials/full_ecModel/data/fluxData.tsv index fd31424d..cae5f8fb 100644 --- a/tutorials/full_ecModel/data/fluxData.tsv +++ b/tutorials/full_ecModel/data/fluxData.tsv @@ -1,2 +1,35 @@ -Condition Ptot grRate glucose (r_1714) CO2 (r_1672) oxygen (r_1992) pyruvate (r_2033) succinate (r_2056) glycerol (r_1808) acetate (r_1634) ethanol (r_1761) -Std 0.4228 0.1 -1.1 2.7 -2.5 8.06E-05 1.00E-05 0.001893379 0.089483471 1.00E-05 +Condition Ptot grRate glucose (r_1714) fructose (r_1709) sucrose (r_2058) xylose (r_1718) CO2 (r_1672) oxygen (r_1992) pyruvate (r_2033) ethyl acetate (r_1765) glycerol (r_1808) acetate (r_1634) ethanol (r_1761) succinate (r_2056) weight source +Std 0.4228 0.1 -1.1 NaN NaN NaN 2.7 -2.5 8.06E-05 NaN 0.001893379 0.089483471 1.00E-05 1.00E-05 0 Standard condition of the protein-constraint step; not used for tuning +glucose NaN 0.36 -13.33333333 0 0 0 NaN NaN 0 0 0.38 NaN 19.82608696 NaN 1 DLKcat +glucose NaN 0.39 -15.44444444 0 0 0 NaN NaN 0 0 0.37 NaN 22.4068323 NaN 1 DLKcat +glucose NaN 0.37 -19.55555556 0 0 0 NaN NaN 0 0 NaN NaN NaN NaN 1 DLKcat +glucose NaN 0.37 -14.8 0 0 0 NaN NaN 0 0 1.1 0.4 22 NaN 1 DLKcat +glucose NaN 0.38 -15.05555556 0 0 0 NaN NaN 0 0 NaN NaN NaN NaN 1 DLKcat +glucose NaN 0.35 -17.66666667 0 0 0 NaN NaN 0 0 0.34 NaN 23.51778656 NaN 1 DLKcat +glucose NaN 0.37 -11.2 0 0 0 19.58 NaN 0 0 0.12 0.43 15.74 NaN 1 DLKcat +glucose NaN 0.37 -13.66666667 0 0 0 NaN NaN 0 0 NaN NaN NaN NaN 1 DLKcat +glucose NaN 0.38 -17.55555556 0 0 0 NaN NaN 0 0 0.41 NaN 28.22463768 NaN 1 DLKcat +glucose NaN 0.32 -25.38888889 0 0 0 NaN NaN 0 0 0.43 NaN 42.73291925 NaN 1 DLKcat +fructose NaN 0.1 0 -6.888888889 0 0 NaN 0 0 0 0.41 NaN 11.14130435 NaN 1 DLKcat +fructose NaN 0.25 0 -12.44444444 0 0 NaN 0 0 0 0.41 NaN 20.256917 NaN 1 DLKcat +sucrose NaN 0.32 0 0 -6.076541046 0 NaN 0 0 0 NaN NaN 19.94202899 NaN 1 DLKcat +glucose NaN 0.025 -0.3 0 0 0 0.75198 -0.73102 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 +glucose NaN 0.05 -0.6 0 0 0 1.3697 -1.3251 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 +glucose NaN 0.1 -1.1 0 0 0 2.4007 -2.3167 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 +glucose NaN 0.15 -1.7 0 0 0 3.6379 -3.5066 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 +glucose NaN 0.2 -2.3 0 0 0 4.8751 -4.6966 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 +glucose NaN 0.25 -2.8 0 0 0 5.906 -5.6881 0 0 0 0 0 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 +glucose NaN 0.28 -3.4 0 0 0 7.0413 -6.6773 0 0 0 0.08 0.11 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 +glucose NaN 0.3 -4.5 0 0 0 8.5667 -6.012 0 0 0 0.41 2.3 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 +glucose NaN 0.35 -8.6 0 0 0 15.2091 -5.3809 0 0 0 0.62 9.5 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 +glucose NaN 0.4 -11.1 0 0 0 19.3207 -5.0171 0 0 0 0.6 13.9 NaN 1 doi:10.1128/AEM.64.11.4226-4233.1998 +glucose NaN 0.4 -20.2 0 0 0 34.4 NaN 0 0 0 1 30 NaN 1 DLKcat +glucose NaN 0.41 -14.22222222 0 0 0 NaN NaN 0 0 0 NaN 25.625 NaN 1 DLKcat +glucose NaN 0.1 -6.44 0 0 0 NaN 0 0 0 1.08 0 9.26 NaN 1 doi:10.1186/1752-0509-2-60 +glucose NaN 0.2449 -10.173 0 0 0 19.109 NaN 0 0 1.2908 0.212 15.494 NaN 1 doi:10.1002/bit.25447 +glucose NaN 0.2287 -13.492 0 0 0 27.31 0 0 0 2.9849 0.328 20.636 NaN 1 doi:10.1002/bit.25447 +glucose NaN 0.1 -6.0184 0 0 0 9.8220288 0 0.0120368 0 1.0351648 0.0240736 8.9734344 0.0270828 1 doi:10.1099/00221287-143-1-203 +glucose NaN 0.2 -11.5927 0 0 0 18.9888426 0 0.0463708 0 2.1098714 0.0695562 17.2499376 0.05216715 1 doi:10.1099/00221287-143-1-203 +glucose NaN 0.3 -17.0704 0 0 0 27.3467808 0 0.1024224 0 3.243376 0.2048448 25.2983328 0.0512112 1 doi:10.1099/00221287-143-1-203 +glucose NaN 0.4 -22.5483 0 0 0 35.3106378 0 0.1803864 0 4.9155294 0.450966 33.6195153 0.0676449 1 doi:10.1099/00221287-143-1-203 +glucose 0.46 0.1 -1.0709285 0 0 0 2.627555 -2.446473 NaN NaN 0.0008575 0 0 NaN 1 doi:10.1016/j.cels.2017.03.003 diff --git a/tutorials/full_ecModel/data/maxGrowth.tsv b/tutorials/full_ecModel/data/maxGrowth.tsv new file mode 100644 index 00000000..8cdb5fa8 --- /dev/null +++ b/tutorials/full_ecModel/data/maxGrowth.tsv @@ -0,0 +1,9 @@ +Condition Ptot grRate glucose (r_1714) fructose (r_1709) sucrose (r_2058) acetate (r_1634) ethanol (r_1761) galactose (r_1710) maltose (r_1931) sucrose (r_2058) mannose (r_1715) weight source +fructose NaN 0.338 NaN -1000 NaN NaN NaN NaN NaN NaN NaN 1 DLKcat +glucose NaN 0.41 -1000 NaN NaN NaN NaN NaN NaN NaN NaN 1 DLKcat +ethanol NaN 0.12 NaN NaN NaN NaN -1000 NaN NaN NaN NaN 1 DLKcat +mannose NaN 0.33 NaN NaN NaN NaN NaN NaN NaN NaN -1000 1 DLKcat +galactose NaN 0.28 NaN NaN NaN NaN NaN -1000 NaN NaN NaN 1 DLKcat +acetate NaN 0.17 NaN NaN NaN -1000 NaN NaN NaN NaN NaN 1 DLKcat +maltose NaN 0.4 NaN NaN NaN NaN NaN NaN -1000 NaN NaN 1 DLKcat +sucrose NaN 0.39 NaN NaN -1000 NaN NaN NaN NaN -1000 NaN 1 DLKcat diff --git a/tutorials/full_ecModel/data/evotuneZeroExch.tsv b/tutorials/full_ecModel/data/zeroFlux.tsv similarity index 100% rename from tutorials/full_ecModel/data/evotuneZeroExch.tsv rename to tutorials/full_ecModel/data/zeroFlux.tsv diff --git a/tutorials/full_ecModel/protocol.m b/tutorials/full_ecModel/protocol.m index 85e46b9d..c49e2741 100644 --- a/tutorials/full_ecModel/protocol.m +++ b/tutorials/full_ecModel/protocol.m @@ -236,8 +236,8 @@ end % Advanced kcat tuning. cmaesKcatTuning fits kcats to measured maximum -% growth rates (data/evotuneMaxGrowth.tsv) and exchange fluxes -% (data/evotuneFluxData.tsv, with data/evotuneZeroExch.tsv), measured on +% growth rates (data/maxGrowth.tsv) and exchange fluxes +% (data/fluxData.tsv, with data/zeroFlux.tsv), measured on % several carbon sources and in aerobic and anaerobic conditions. It scores % each set of kcats by the carbon-weighted root-mean-square error (RMSE) % against these data and searches with CMA-ES, an evolutionary optimiser. @@ -287,7 +287,7 @@ % All usage reactions draw from the protein pool, also those with a % measured bound. The protein pool of the condition uses the Ptot of the % condition from data/fluxData.tsv and the f of the proteomics data. -fluxData = loadFluxData(); +fluxData = loadFluxData(fullfile(params.path,'data','fluxData.tsv')); f = calculateFfactor(ecModel,'protData',protData); ecModel = setProtPoolSize(ecModel,'Ptot',fluxData.Ptot(1),'f',f,'sigma',sigma);