Version Compatibility

Reference examples tested with: mebocost 1.0+, scanpy 1.10+, anndata 0.10+

Before using code patterns, verify installed versions match. If versions differ:

• Python: pip show <package> then help(module.function) to check signatures

If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API rather than retrying.

Metabolite-Mediated Cell Communication

"Find which cell types exchange metabolites" -> Infer a metabolite's presence from the cells expressing its synthesizing enzymes, then score communication to cells expressing its sensor or transporter.

• Python: mebocost.create_obj() -> infer_commu() (metabolite-sensor scoring)

Governing Principle

Metabolite-mediated communication is a DOUBLE inference and the most speculative layer of cell-cell communication. scRNA-seq never measures metabolites; their levels are inferred from the expression of synthesizing enzymes, then a chain of further assumptions is stacked: enzyme mRNA -> enzyme protein -> enzyme ACTIVITY -> metabolic FLUX -> intracellular metabolite POOL -> SECRETION/export -> extracellular concentration in space -> import/SENSING by the receiver. Every arrow is an assumption and none is measured. On top of this sit all the ligand-receptor caveats (proxy, no spatial geometry in dissociated data, abundance and depth confounds, ambient RNA), so the output is hypothesis-generation ONLY. A defensible claim is never "cell A produces metabolite X" but "cell A expresses the machinery consistent with producing X". Validation is non-optional and means metabolomics (LC-MS), mass-spectrometry imaging (MALDI/DESI), isotope tracing, or perturbation of the enzyme or sensor - not another expression-based method. State the enzyme->flux->level->sensing chain explicitly whenever reporting a result.