Create case studies

[jonrkarr]

Calibrate or re-calibrate a kinetic model of a mammalian signaling pathway

First, find an appropriate model from the curated portion of BioModels

• Represents Homo sapiens or Mus musculus
• Represents a signaling pathway
• Species are annotated with UniProt ids

Here's the SBML files for the curated portion of BioModels.

Second, try to use the PAXdb data either to

• (a) Calibrate the initial species concentrations (i.e., predict the values in the SBML file; the values in the SBML file would serve as the "gold" standard)
• (b) Re-calibrate the initial species concentrations to represent a different organism or cell type

Enhanced Yeast FBA model

Any version could be used as a starting point.

• https://github.com/SysBioChalmers/yeast-GEM
• v8: https://doi.org/10.1038/s41467-019-11581-3
• v7: https://doi.org/10.1089/ind.2013.0013
• v6: https://doi.org/10.1093/database/bat059
• v5: https://doi.org/10.1186/1752-0509-6-55
• v4: https://doi.org/10.1186/1752-0509-4-145
• v3: There is no model labeled v3
• v2: There is no model labeled v2
• v1: https://doi.org/10.1038/nbt1492

Ways to improve the yeast model

• Additional flux constraints equal to {k_cat} * {enzyme abundance}
• More detailed biomass reconstruction using metabolite concentrations from YMBD

Potential ways to demonstrate improvement of the FBA model

• Increase in the number of reactions with non-zero predicted flux
• Increase in the number of exchanged metabolites (metabolites imported from the extracellular media)
• Improved prediction of gene essentiality (would be due to more detailed biomass)
• Resolution of any problem with the original model described in one of the original publications
• Prediction of distribution of growth rates of single cells (demonstrates improvement in capabilities rather than in biological validity)