Roadmap for the next PEtab version

[dilpath]

• move repo to PEtab-dev org
• rename repo benchmark-problems (i.e. PEtab-dev/benchmark-problems)
• rename Benchmark-Models directory to problem (i.e. PEtab-dev/benchmark-problems/problems)
• add metadata to each problem in its PEtab YAML file (closes Automatic generation of general information file #25)
  • create schema for such metadata
  • include objective function value evaluated at nominal parameters (possibly closes Be careful with comparison of objective function values #17 unless that it about curation itself). also optionally the Chi2.
  • include DOI to original publication
  • copy all relevant metadata from model to PEtab YAML, including e.g. model ID (closes Change <model> metaid in all models #155)
• create script to automatically extract interesting problem "features" then produce a "problem" vs. "feature" matrix TSV file for convenience (closes Create table of benchmark models and their model/data peculiarities #18 Write function to generate benchmark_model_characteristics.xlsx automatically #19). example features (maybe difficult to extract): "splines", "log-scale observables", "parameter-dependent initial conditions". Include all features currently in README overview table.
• add check for correctly-formatted simulation tables (closes Update simulatedData files #20)
• update all PEtab files (e.g. PEtab versions) (closes Add PEtab version info to files #1)
• update contribution guide (closes Add contribution guide #32)
• auto-update CITATION.cff to match Zenodo on new "releases" (closes Add CITATION.cff #164)
• consolidate scripts and Python package (Machine readable model property overview #169)
• require(?) PEtab Results with each problem to ensure "exact"/"easy" reproducibility
• add "Support" column to overview table, that links to each model's respective issue, where users can request model-specific support

[dweindl]

• include objective function value evaluated at nominal parameters

That would be super helpful. Also the gradients.
That will make this collection more relevant for regular testing, not just benchmarking.