Data and code for two studies of the phenibut GABA-B axis: alchemical transfer free energies, and KCTD-subunit genetics with FAERS pharmacovigilance

Supporting data and code for two manuscripts by Andrei Mihai Visalon (University of Bucharest). Paper A (computational): alchemical transfer method (ATM/RBFE) free energies for the phenibut enantiomers at the GABA-B receptor, licensed on a two-sided validation edge and reported as three independent seeds. The claim is the sign — R-phenibut is the eutomer — and the magnitude is explicitly declined. The same protocol is then shown to fail at a second target, α2δ-1, and that failure is reported in full. Paper B (genetics and pharmacovigilance): the GABA-B auxiliary-subunit axis, a cis-Mendelian-randomization instrument for KCTD16 that is significant in FinnGen but does not replicate outside Finland (Howard 2019 OR 1.02, p = 0.34; colocalization collapsing from PP.H4 0.952 to 0.0014), and a reporting-bias-aware FAERS disproportionality analysis. This archive is assembled so the published free energies can be re-derived from it alone, and that is verified rather than asserted. scripts/verify_deposit_selfsufficient.py runs the study's own UWHAM estimator over atm_drains/ and reproduces all three seeds' per-leg free energies and ΔΔG to within 0.0004 kcal/mol of the published values. Contents: the single-source-of-truth results record; all analysis and figure-generation scripts; per-replica UWHAM inputs for the three enantiomer seeds plus the serialized OpenMM System the force-field reading is made against; receipt directories for every analysis (MR robustness panel, both colocalizations, the contact-occupancy recomputation, the 8IF3-to-UniProt residue mapping, the Dambrova figure-axis read); all manuscript figures at 300 dpi or better; and a SHA-256 manifest covering every file. Raw molecular-dynamics trajectories (~1.7 GB of .xtc and checkpoint files) are not included: they are needed to re-run the dynamics, not to reproduce any reported number. They are available on request. Scope. Both manuscripts are mechanism and hypothesis framing only. Neither contains, and neither should be read as containing, dosing, titration, tapering, substitution, sourcing or any other human-use guidance. Phenibut is a psychoactive substance associated with dependence and withdrawal; nothing here addresses its use in humans.