Integration of utility from SPECIMEN into refinegems

[cb-Hades]

Tracker for the integration of the utility functions and ideas from SPECIMEN into refinegems.

Possible Problems:

• changed load_model_cobra to be able to read in all types of files -> 🟥 WARNING: requires python 3.10 (is related to issue Implement parsing of different formats #44)

TODOs:

• compare BOF / normalisation functions
• extend medium module - compare with growth ?, possible relation to issue Update the growth module #104
• 🔴 compare EGC / futile cycles functions -> Enhance implementation of EGC algorithm #114
• analysis: SBO / pathways / statistical - overlaps, try to combine and extend. report here? (compare with investigate?)
• downloading BiGG / MetaNetX (?) - rest of setup only relevant for SPECIMEN, only check for ideas for refinegems
• 🔴 set-up module and other utilities? -> new issue Evaluate & add entry points #118

[cb-Hades]

Progress for medium module (added functions for ...):

• ex/import from/to cobra models
  • currently only BiGG namespace based possible, but options for extention
• aerobic/anaerobic check + set
• combine media
• set default flux
• direct export from a medium object
  • to file
    • 🟪 IDEA: add export to pdf to get a nice info - maybe even better to get rst from docs directly
  • to cobra -> 🟧 TODO: currently only implemented for BiGG namespace, but options for more -> Modifiy modules to work with models not using BiGG identifiers #36

[cb-Hades]

Progress for util:

• function for reading in cobra models added to io
• function for reannotating SBO for "fixing" memote score

[cb-Hades]

Note to simulation / plotting:

• plot / report function and classes for growth simulation have been partially integrated during the growth module overhaul Update the growth module #104
• added the auxotrophy simulation function to refinegems
• added the KEGG pathway analysis function to the pathways module and the corresponding report class to the reports module
  • adjusted docs and colour scheme to fit the rest of the modules (as good as possible as pie charts cannot take a gradient as input)