Notes 15..01.05 meeting notes

15..01.05 meeting notes

• Get all SBRG models from Ali’s model collection
• set up a pre-loading script to analyze SBML models (Maybe a week to get this running - Zak will head this)
  • load models into COBRApy
  • try to compute
    • make a COMPUTABLE flag in the models table (some will not compute by default, but will be served with medium to make them computable… notes would include all possible media and objectives, and links to papers describing setup)
  • make sure there is only one objective
  • check that IDs match https://github.com/SBRG/BIGG2/wiki/BIGG2-ID-Proposal-and-Specification (e.g., biomass should start with biomass_ and exchange rxns are EX_ and demand are DM_… all other rxns must be mass balanced)
• Come up with a more solid plan for comparing metabolites (make sure all metabolites align between models) (Jon will head this with Philip. If the MetaNetX team has notes, then we can make the mapping quick. If not, then we’re looking at maybe moving forward without this)
  • use a combination of BIGG IDs and MetaNetX (let’s email them about aligning the remaining 1200 metabolites)
  • waiting for a followup from Jon
• Try to match reactions based on both metabolites and BiGG IDs, then flag the differences (Justin will be in charge of this)
  • also look for linear combinations of reactions in the database
  • hash the metabolites and coefficients
• Be sure to record the previous metabolite and reaction IDs somewhere (old_id_synonyms tables) (Justin)
  • Make it possible to convert analysis with old BIGG IDs to the new BIGG2 IDs
  • Does not need to be visible on the website, but we could provide and API for converting IDs
  • for reactions, models, and genes
  • deal with the “deleted” reactions
• for genes (aligning all genes to an NCBI ID) (Justin)
  • flag all the genes that are not in the genome annotation
    • Count them up
  • always use locus ids as the gene IDs
  • use NCBI gene names as the gene names (readable gene name)
  • fix any models that do not conform to this (including the gene_reaction_rules)
  • keep s0001 in the database for now, but flag it as not a real gene
• Methods to quantify our success (internal, paper) (Zak and Justin… this is still unclear)
  • Matched x% of (metabs, rxns) iAF1260 content in iJO1366
  • % of genes from model found in NCBI genome annotation
• use postgres database dump/load for persisting the database state (Justin)
• in the BIGG website, remove all placeholders and fill in missing pages/dead links (Zak and Justin)
  • ask Zak where appropriate
  • on subpages (i.e. not the homepage) replace the SBRG title and logo with the BIGG title
  • http://equilibrator.weizmann.ac.il/ as an example of UI choices
  • capitalize Universal
• Database and interface QA and checking (Philip)
• https://www.dropbox.com/s/rnf5pohw2rgcrgn/15..01.05%20BiGG%202%20whiteboard%20mtg.pdf?dl=0