GECKO 2.0.3

• Refactor:
  • speed improvements around matchKcats and getEnzymesCodes (PR #154)

GECKO 2.0.2

• Features:
  • FSEOF implementation for ecModels in /geckomat/utilities/ (PR #119).
  • Routines for quickly generating context-specific ecModels from a general ecModel, also in /geckomat/utilities/ (PR #120).
• Fixes:
  • Closes #116: Migrated CI to Travis-CI.com (PR #118).
  • prot_abundance.txt can now be an empty file (for organisms not present in pax-db) without the pipeline erroring (PR #124).
  • Removes storage of metabolite notes in the ecModel, which were causing an error in COBRA (PR #125).
• Documentation/Styling/Others:
  • Added contributing guidelines, code of conduct and PR/issue templates (PR #115).
  • Closes #123: Simplified GECKO pipeline log (PR #128).

GECKO 2.0.1

• Fixes:
  • UBs are only changed for strictly required proteins (PR #103).
  • Closes #101: All yeast models in the toolbox are built on the same yeast-GEM version 8.1.3 (PR #105).
  • Solved bug in generate_protModels.m that constrained both the biomass and growth reactions (PR #107).
  • Updated env. list after release of pandas 1.0 (PR #110).
  • A folder for ecModels is added in case it doesn't exist already (PR #111).
  • Closes #112: Avoid using the variable name version for potential conflicts with Matlab (PR #113).
• Style:
  • Closes #102: All UBs in repo changed from +Inf to +1000 (PR #109).

GECKO 2.0.0

• Features:
  • A main utility for constraining an ecModel with proteomics data stored in data files, and ensuring the model can grow at the provided conditions (PR #82).
  • More constraining options in FVA utility (PR #95).
  • An option for reducing display of fitGAM.m (PR #82).
  • A chemostat simulation utility (PR #82).
• Fixes:
  • Fixes #73: Bug in getKcat.m (PR #75).
  • Fixes #78: Updated BRENDA URL (PR #91).
  • FVA utility: Fixed directionality bugs in MAX_min_Optimizer.m (PRs #69 & #95) & optimization bug in comparativeFVA.m (PR #95).
  • geckopy: Updated csv reader in geckopy test (PR #83) and met id parsing (PR #93).
  • Fixed subfolder bug in modifyKcats.m (PR #80).
• Refactoring:
  • Introduced getModelParameters.m for defining all input parameters, for streamlining the main ecModel generation pipeline (PR #76).
  • Simplified constraining procedure in the FVA utility (PR #95).
• Documentation:
  • Improved documentation of several main functions (PR #76).
  • Updated Matlab requirements (PR #96).

GECKO v1.3.5

• Features:
  • Additional options for output tables from modifyKcats.m & topUsedEnzymes.m (PR #61)
  • keggID is now an input for updateDatabases.m (PR #62)
  • Backwards compatibility with any yeastGEM from 8.0.0 onwards (PR #66)
  • New utilities:
    • getSubset_ecModel.m, for getting context-specific ecModels (PR #64)
    • getKcat.m, for retrieving kcats (PR #67)
• Fixes:
  • Fixed bug in aconitase kcat & misc. error handling (PR #62)
• Refactoring:
  • Speed improvements in topUsedEnzymes.m (PR #61)
  • Reduced display of several functions (PR #62)
  • Simplified changeMedia_batch.m and made more generic constrainEnzymes.m & flexibilizeProteins.m (PR #63)
• Style:
  • Changed EOL to LF (unix default) (PR #68)
• Documentation:
  • Documented input/output of topUsedEnzymes.m & truncateValues.m (PR #61)
  • Added/updated documentation of changeMedia_batch.m, constrainEnzymes.m, flexibilizeProteins.m & getConstrainedModel.m (PR #63)

GECKO v1.3.4

• Features:
  • Generalization of measureAbundance.m to receive any PaxDB file, a relative proteomics dataset, or even nothing at all (PR #58).
  • New utility: Comparative FVA between a model and its enzyme-constrained version (PR #57).
• Fixes:
  • Consistent definition of what data is in uniprot.tab (PR #48).
  • Proper use of measureAbundance.m from within constrainEnzymes.m (PR #56).
• Refactoring:
  • Switch all functions that add/change rxns/genes from COBRA to RAVEN (PR #48).
  • Avoid any functions from Simulink (PR #48).

GECKO v1.3.3

• Fixes:
  • Fixes #15: Binary results from the model (ecModel.mat, ecModel_batch.mat & enzData.mat) are no longer stored in repo (PR #52).
  • Misc. fixes in the biomass composition + GAM calculations (PR #53).
• Refactoring:
  • Speed improvement in misc. functions (PR #49).
  • Added sumProtein.m for easier use when creating new ecModels (PR #53).
• Documentation:
  • Documented better which scripts/data should be changed and which are optional when adapting geckomat to produce a new ecModel (PR #53).

GECKO v1.3.2

• Features:
  • Name & version of the model are now read/stored from/as model fields (PR #42).
  • Pipeline now works for any objective function (PR #47).
• Fixes:
  • Fixed bug from #39 that saved the .mat file with the wrong name (PR #42).
  • Adapted pipeline to deal with multiple gene IDs for 1 protein / multiple protein IDs for 1 gene, for dealing with human-based GEMs (PR #43).
  • changeMedia_batch.m modified to reflect the Y6 minimal media composition (PR #47).
• Refactoring:
  • Performance improvements to getConstrainedModel.m and sigmaFitter.m (PR #47).
  • fitGAM.m is now only called from inside scaleBioMass.m (PR #47).

GECKO v1.3.1

• Features:
  • Adapted the pipeline to work with yeast-GEM, including loading, processing and saving the model. Current model is constructed from yeast v8.1.3 (PR #39).
  • When constructing ecModel_batch, lipid fraction is now scaled together with protein and carbohydrate fractions (PR #39).
• Fixes:
  • geckopy tests flexibilized to comply with yeast-GEM (PR #39).
• Refactoring:
  • Reorganized the repo, making a division between geckomat (Matlab part for generation + simulation of ecModels) and geckopy (Python part for simulations of ecYeastGEM) (PR #40).
  • Parameters f (mass fraction of enzymes in model), Pbase, Cbase, Lbase (biomass composition) and GAM (growth-associated ATP maintenance) are now automatically computed (PR #39).
  • Added RAVEN as a dependency for geckomat (PR #38).
  • Changed most COBRA functions in pipeline to RAVEN functions (PR #39).

GECKO v1.3.0

• Features:
  • Protein flexibilization: When proteomic measurements are provided, individual protein levels will now be iteratively flexibilized by the pipeline if the model results to be overconstrained, based on a provided growth rate. After this, flexibilized protein exchange pseudoreaction upper bounds will be set to the their flux values from a parsimonious FBA simulation (PR #34).
  • Utilities: Included a folder with useful functions (PR #34).
• Fixes:
  • Fixes #14: CI is no longer failing, as model location, model naming and metabolite ID naming were corrected. test_adjust_pool_bounds was simplified to test with only 1 essential protein (PR #28).