New

checkForMissingStereo(model, trainingModel)

USAGE

missingStereo = checkForMissingStereo (model, trainingModel)

INPUTS

• model – structure with fields:

  • .mets

  • .inchi.standard

  • .inchi.standardWithStereo

• trainingModel – structure with fields:

  *.inchi.standard *.inchi.standardWithStereo

OUTPUTS

missingStereo

createTrainingModel(trainingModel, trainingMolFileDir, forceMolReplacement, printLevel)

create the training model, or update it with additional data

OPTIONAL INPUTS: trainingModel: molFileDir: directory of the mol files forceMolReplacement: force the replacement of the existing mol files

with newly acquired ones

printLevel:

OUTPUTS

trainingModel – trainingModel structure with following additional fields: * .mets m x 1 metabolite abbreviations * .rxns n x 1 reaction abbreviations * .metKEGGID m x 1 trainingModel.cids; * .metChEBIID m x 1 ChEBI identifier of the metabolite. * .inchi - Structure containing four m x 1 cell array’s of

IUPAC InChI strings for metabolites, with varying levels of structural detail.

• .inchi.standard: m x 1 cell array of standard inchi

• .inchi.standardWithStereo: m x 1 cell array of standard inchi with stereo

• .inchi.standardWithStereoAndCharge: m x 1 cell array of standard inchi with stereo and charge

• .inchi.nonstandard: m x 1 cell array of non-standard inchi

• .inchiBool m x 1 true if inchi exists

• .molBool m x 1 true if mol file exists

• .compositeInchiBool m x 1 true if inchi is composite

driver_addChemoinfoToTrainingModel

Generate database

driver_createTrainingModel

Cretate a validated training model for thermochemical estimation Author: Ronan Fleming, German Preciat, Leiden University Reviewers: INTRODUCTION

PROCEDURE Configure the environment All the installation instructions are in a separate .md file named vonBertalanffy.md in docs/source/installation

With all dependencies installed correctly, we configure our environment, verfy all dependencies, and add required fields and directories to the matlab path.

loadTrainingData(param)

Generates the structure that contains all the training data needed for Component Contribution.

USAGE

trainingModel = loadTrainingData (formation_weight)

INPUT

formation_weight – the relative weight to give the formation energies (Alberty’s data) compared to the reaction measurements (TECRDB)

OUTPUT

trainingModel – structure with data for Component Contribution *.S m x n stoichiometric matrix of training data *.cids: m x 1 compound ids *.dG0_prime: n x 1 *.T: n x 1 *.I: n x 1 *.pH: n x 1 *.pMg: n x 1 *.weights: n x 1 *.balance: n x 1 *.cids_that_dont_decompose: k x 1 ids of compounds that do not decompose

prepareTrainingData(model, printLevel, params)

Given a standard COBRA model, adds thermodynamic data to it using the Component Contribution method

USAGE

training_data = prepareTrainingData (model, printLevel, params)

INPUT

model – COBRA structure

OPTIONAL INPUTS

• printLevel – verbose level, default = 0

• params.use_cached_kegg_inchis

• params.use_model_pKas_by_default

• params.uf – maximum uncertainty

OUTPUTS

trainingData – structure with the following fields:

• .S - the stoichiometric matrix of measured reactions

• .G - the group incidence matrix

• .dG0 - the observation vector (standard Gibbs energy of reactions)

• .weights - the weight vector for each reaction in S

• .Model2TrainingMap

reverseTransformTrainingData(trainingModel, use_model_pKas_by_default, model)

Calculate the reverse transform for all reactions in trainingModel.

INPUT trainingModel

OPTIONAL INPUT use_model_pKas_by_default model

OUTPUT trainingModel.DrG0: n x 1 standard Gibbs energy