State a falsifiable physiological question
Every program defines the dysregulated state, the useful function to preserve and the therapeutic response being investigated. The hypothesis must identify an observation that would challenge it.
Design from the required behaviour
Target, modality and molecular design are chosen in relation to the required pharmacology. State, timing, signalling, endogenous demand and functional consequence are treated as explicit design variables when the biology makes them relevant.
Challenge models before advancing them
Literature claims are checked against their sources. Computational methods are tested against appropriate controls, alternative explanations and predefined failure criteria. Reproducibility and method fitness are assessed before an output informs a program decision.
Translate predictions into experiments
Computational work characterizes targets, generates molecular hypotheses and prioritizes direct tests. Synthesis and experimental pharmacology provide the next evidence for molecular and functional claims.
Follow the result
Results can support the current design, expose a different pharmacological direction or end the line of work. Aeviant records those decisions and updates the program when the evidence changes.