Integrated Quantitative Systems Pharmacology model of menopausal hot flushes and night sweats, linking ovarian follicle depletion and loss of estrogen negative feedback to arcuate KNDy-neuron (kisspeptin/neurokinin B/dynorphin) hypertrophy, NKB→NK3R-driven narrowing of the hypothalamic thermoneutral zone, and the resulting hot-flush frequency/severity — together with the modern pharmacology stack (NK3R antagonist fezolinetant, dual NK1/NK3 antagonist elinzanetant, hormone therapy, SSRIs/SNRIs, gabapentin, clonidine, oxybutynin) and downstream organ- system consequences (sleep, mood, bone, lipids, genitourinary syndrome).

Vasomotor symptoms (VMS) — hot flushes and night sweats — affect up to 80% of women during the menopause transition and persist a median of 7.4 years (SWAN). Declining ovarian follicle numbers reduce estradiol (E2) and inhibin B, removing negative feedback on the hypothalamus. Arcuate (infundibular) KNDy neurons, which co-express kisspeptin, neurokinin B (NKB), and dynorphin, undergo estrogen-withdrawal-driven hypertrophy and increased NKB output. NKB acts on NK3R both autosynaptically (a feed-forward loop) and on neurons of the median preoptic nucleus (MnPO) thermoregulatory center, narrowing the thermoneutral zone so that small elevations in core temperature trigger inappropriate heat-dissipation responses (cutaneous vasodilation, sweating) — the hot flush. Downstream consequences include sleep fragmentation, mood disturbance, accelerated bone loss, adverse lipid changes, and genitourinary syndrome of menopause. The 2023 approval of fezolinetant, a selective NK3R antagonist, and the late-stage dual NK1/NK3 antagonist elinzanetant directly target this mechanism, complementing hormone therapy and older nonhormonal options (SSRIs/SNRIs, gabapentin, clonidine, oxybutynin).

1. Ovarian aging & follicle depletion (STRAW+10 staging)
2. HPG-axis feedback loss (GnRH pulse generator → FSH/LH → E2/E1)
3. Arcuate KNDy-neuron biology (Kiss1 · NKB/TAC3 · dynorphin · ERα brake)
4. NKB→NK3R signaling & median preoptic (MnPO) thermoregulatory center
5. Sympathetic effector arm (cutaneous vasodilation, sweating, flushing)
6. Clinical VMS phenotype (hot flush trigger → frequency → severity → night sweats)
7. Sleep architecture & CNS/mood consequences
8. Skeletal consequences (RANKL/OPG · osteoclast · CTX/P1NP · BMD · fracture)
9. Cardiometabolic & lipid consequences (LDL/HDL, endothelial function, adiposity)
10. Genitourinary syndrome of menopause (vaginal atrophy, pH, dyspareunia)
11. Drug PK/PD — NK3R antagonist (fezolinetant) & dual NK1/3 (elinzanetant)
12. Drug PK/PD — hormone therapy (oral/transdermal estradiol, CEE, progestogen)
13. Drug PK/PD — nonhormonal agents (SSRI/SNRI, gabapentinoid, α2-agonist)
14. Clinical endpoints, biomarkers & PROs (HF frequency/severity, MENQOL, PSQI, DXA)

• Drug PK (5 drug classes, 17 compartments) — fezolinetant, elinzanetant, oral estradiol, transdermal E2 patch, paroxetine, venlafaxine + active metabolite ODV, gabapentin (saturable absorption), clonidine (literature PK anchors in vms_references.md).
• Ovarian/HPG axis (3 compartments) — follicle pool, plasma E2, FSH — driving the estrogen-withdrawal signal.
• KNDy/thermoregulatory network (2 compartments) — KNDy/NKB tone and thermoneutral-zone (TNZ) half-width, with NK3R-antagonist and nonhormonal-drug reversal terms feeding directly into TNZ narrowing.
• Clinical readouts (6 compartments) — hot-flush frequency, hot-flush severity, PSQI sleep score, mood composite, bone resorption (CTX) and BMD, LDL-C.

1. Patient profile — adjustable age/BMI/menopausal stage/severity & baselines.
2. Drug PK — log-scale plasma concentrations for the nine tracked drug moieties.
3. HPG / KNDy axis — follicle pool, E2, FSH, and KNDy/NKB tone trajectories.
4. Thermoregulation — thermoneutral-zone (TNZ) half-width vs. premenopausal reference.
5. Hot-flush endpoint — frequency (events/day) and severity (0-3).
6. Sleep / Mood — PSQI sleep-quality index and mood composite.
7. Scenario comparison — runs all 10 scenarios with the chosen profile.
8. Biomarkers / safety — bone (CTX, BMD), LDL-C, endpoint summary table.

# 1) Render the mechanistic map
dot -Tsvg vms_qsp_model.dot -o vms_qsp_model.svg
dot -Tpng -Gdpi=150 vms_qsp_model.dot -o vms_qsp_model.png

# 2) Simulate scenarios in R
install.packages(c("mrgsolve","dplyr","tidyr","ggplot2","shiny",
                   "shinydashboard","DT"))
library(mrgsolve)
mod <- mread_cache("vms_mrgsolve_model.R")
out <- mrgsim(mod, ev(amt=45, cmt="GUT_FEZ", ii=1, addl=168), end=168)
plot(out, c("HF_freq_perday","TNZ_degC","E2_pgmL","PSQI_sleep"))

# 3) Launch the dashboard
shiny::runApp("vms_shiny_app.R")

• Designed for research, education, and hypothesis generation — not clinical decision support.
• All PK/PD parameters are population means; inter-individual variability needs to be added with omega() blocks for population simulations.
• The hot-flush "frequency" compartment is a continuous surrogate for what is physiologically a discrete, Poisson-like event process; it captures trends in mean daily rate, not individual event timing.
• Hormone-therapy safety outcomes (VTE, breast cancer, stroke risk from WHI) are referenced in the bibliography but intentionally not modeled as ODE compartments — this model focuses on VMS efficacy and the bone/lipid/GSM mechanistic axis.

Inherits the repository LICENSE.