I think I have figured out at least one specific phenotype of PFS, and it is different from the "allopregnanolone" theory

Powers' key genetics-first post: he proposed a specific PFS phenotype whose "base, core defect" is defective UGT2B17, disabling testosterone's main glucuronidation exit pathway. Carriers show shockingly low 3α-androstanediol glucuronide and near-zero urinary testosterone on DUTCH despite normal serum T; finasteride then closes the remaining DHT exit, causing intracellular androgen overcrowding, receptor downregulation, and eventual epigenetic silencing of androgen signaling. Minor defects in UGT2B15/UGT2B7, weakening SULT2A1 variants, HSD17B2 defects, and defects across AKR1C1–C4 and SRD5A1 (epigenetically silenceable) compound the phenotype. A January 2026 edit added SLCO1B1 (rs200994482, c.1865+1G>A het, likely pathogenic — Rotor syndrome) as producing the same phenotype via defective glucuronidated-steroid recycling. This post motivated the SIDEfxHUB community data-collection effort (morning hormone panel + androstanediol glucuronide blood test + DUTCH test). Full gene-level detail in powers-gene-list.md.