Association is not causation
Observational or correlational relationships should not be described as causal without appropriate perturbation, mediation, transfer, rescue, or dependency evidence.
Evidence & Mechanism is the methodological signature of the site: a framework for aligning the strength of a scientific claim with the strength and type of evidence supporting it.
Evidence pathway
Observational or correlational relationships should not be described as causal without appropriate perturbation, mediation, transfer, rescue, or dependency evidence.
Changes in expression, abundance, or localization may support a mechanistic hypothesis but do not alone establish functional dependency.
Network pharmacology, docking, molecular dynamics, enrichment analysis, computational target prediction, and multi-omics integration are hypothesis-generating or supportive unless experimentally validated.
A change in a metabolite, microbial taxon, receptor, or pathway marker does not establish that it mediates the observed phenotype.
Biological independence, replication structure, donor/animal/sample hierarchy, and pseudoreplication can determine whether an apparent effect is statistically interpretable.
Language should distinguish association, modulation, involvement, dependency, necessity, sufficiency, rescue, and causation.
The goal is not to make every conclusion stronger. It is to make every conclusion proportionate to the evidence.