Maya-Vaidya P9: Viparyaya The Sign of the GABAergic Neuron–Glioma Synapse Is Under-Determined by Existing Data: A Spiking-Network Map of the Growth/Suppression Boundary

Version 2 (2026-08-04). Revised to the NLL Universal Paper Format v6. The original text is retained in full; nothing has been deleted. Corrections appear as marked blocks placed at the section that carries the claim, and each one states what the paper said, what the data show, the corrected claim, why it happened, and what still stands. This version adds one narrow disclosure; nothing is retracted.§4.6 already discloses that the r = 1.000000 boundary is an algebraic identity, but exempts the zero-crossing location. That exemption does not hold: the crossing is set by f_syn and the AMPA drive, both tagged INFERRED in the paper's own provenance table, so the 3.28 mV displacement and the ±1.5 mV precision asked of experimentalists inherit the same dependency — slope, displacement and ask are one algebraic step apart, not three findings.This paper is the reference implementation for the corpus standard: physical units, calibrated firing rates, a full provenance table, matched controls, a disclosed discarded run, a deferred hypothesis, and a reproduction test. The five-pair classification is unaffected. Everything below this line is the original description from version 1. It is retained unchanged for the record. Where it conflicts with the corrections above, the corrections stand. Two framings it repeats have since been withdrawn in full — the Bhaya Quiescence Law and the Buddhi S-Curve. Both are addressed in Maya-Meta P1, now superseded, and in the self-audit of Maya-Meta P2. Maya-Vaidya P9: Viparyaya (विपर्यय; inversion — mistaking a thing for its opposite) asks a question the published literature cannot currently answer: does a GABAergic neuron→glioma synapse feed a tumour or brake it? GABA carries no intrinsic sign. Opening a GABAA receptor opens a chloride conductance, and whether chloride enters or leaves is set by the chloride reversal potential ECl against the cell’s operating potential. Glioma cells accumulate chloride to roughly threefold the neuronal concentration, which can invert the effect entirely. The problem is the number: reported intracellular chloride in glioma spans a tenfold range across laboratories on comparable tissue — 13 mM (Barron 2025, IDH-wildtype), 51–60 mM (Barron 2025, DMG), 100–105 mM (Habela 2009, gramicidin perforated patch), 140 mM (Sontheimer, personal communication). A tenfold spread in the quantity that sets the driving force means the sign of the effect is not determined by the published data. Rather than select a value and report the consequence — a choice presented as a result — this work sweeps it. The substrate is the P7/P8 leaky integrate-and-fire circuit (800 excitatory / 200 inhibitory) driving 150 non-spiking glioma units through three measured channels: sparse AMPA onto 15% (P7, DOI 10.5281/zenodo.21489850), a diffusible paracrine field reaching all of them (P8, DOI 10.5281/zenodo.21504243), and — new here — a GABAergic synapse onto 40%. 1,236 simulations, 20 published seeds, pure NumPy/SciPy on CPU. (1) The sign reverses inside the measured span. Holding Vrest at −30.1 mV and sweeping ECl from −61.3 to −6.82 mV, the effect on modelled proliferation crosses zero at ECl = −26.61 mV; below it GABA is a net brake, saturating on a floor at 0.441416 where only the paracrine channel survives. (2) The boundary is a straight line, and not the obvious one. Over the two-dimensional (Vrest, ECl) plane the growth/suppression contour is ECl* = 0.891 × Vrest (Pearson r = 1.000000, n = 7 rows), not the naive identity ECl = Vrest. (3) The displacement is made by glutamate, not GABA. Removing the AMPA channel returns the boundary to identity (−30.00 mV against Vrest = −30.1 mV) — tonic glutamatergic drive holds the operating point above rest, and it is the operating potential the reversal competes against. This was not the pre-registered mechanism: H2 predicted the GABAergic shunt, and the prediction held for the wrong reason. (4) The central result: the five published parameter pairs fall on both sides of the boundary — 3 growth-promoting, 2 suppressive. A published disagreement about the sign need not be a contradiction; it is what a boundary looks like when different preparations sit on different sides of it. (5) The honest limit, stated in Results rather than buried in Limitations: the displacement is 3.28 mV, smaller than the standard deviation on every published EGABA measurement (±3.7, ±4.9, ±7.9 mV), and it reclassifies 0 of the 5 pairs. It is offered as a prediction with a stated precision requirement (≈ ±1.5 mV on the same cell), not as a usable correction. (6) Along the cell cycle the effect turns at φ = 0.548 while the driving force turns later at φ = 0.593. Reported against itself. Kill conditions were pre-registered and are published with their verdicts, including the rows that did not go the author’s way. H4 — DEFERRED: the shunt-dominated regime that would most cleanly reconcile the growth and suppression literatures could not be tested, because drive and shunt are two aspects of one conductance and a time-averaged readout is insensitive to a pure shunt; testing it needs a fluctuating-input regime, which is a redesign, not a patch. It is not falsified — it is unexamined, and saying so is the report. H5 — PASS, but never at risk: Barron’s measured −32.2 mV bumetanide shift lands at ECl = −57.2 mV, inside the span already swept, and the two agree to 4 × 10−7 — so it is a placement of a measured, untuned shift on the map, not an independent validation, and is reported as such. A dedicated Results subsection, “What is arithmetic, not biology”, names every part of the output fixed by construction. One clean pre-registered null: the GABAergic and paracrine channels are independent (0.004815 with and without, identical at every reported digit). No biological novelty is claimed: every mechanism modelled is somebody else’s laboratory work (Ullrich & Sontheimer 1997; Habela & Sontheimer 2007; Habela 2008, 2009; Barron/Monje 2025). The contribution is the map and the under-determination argument. Confounds stated plainly: combining Vrest from GBM lines with EGABA from DMG lines is an INFERRED modelling step — no laboratory has measured both in the same cell at the same cycle phase, and that absence is the paper’s subject rather than a gap papered over; the linear ramp between floor and ceiling is arithmetic, not biology; one chloride efflux route is modelled where Habela 2008 names two; the post-mitotic NKCC1 influx limb rests on personal communication and is tagged PERSCOMM throughout; glioma units neither divide, migrate nor die, and cell-cycle state is a parameter trajectory, not a simulated process; units are arbitrary with no external calibration. Scope-lock per Maya-Meta P2 (DOI 10.5281/zenodo.20621503): reproducible signatures within THIS architecture under THIS protocol — Property, not Law. What would change the author’s mind, written into the paper: a paired recording of EGABA and membrane potential in the same glioma cell, under two levels of synaptic drive, showing no change in the sign of the GABA effect. That measurement has not been made, and naming it is the paper’s most useful sentence for an experimentalist. Reproducibility: all 20 seeds, raw per-seed arrays, the arms matrix with every pairwise Mann–Whitney U test, and a 15-test suite that re-derives every published number from the raw data ship with the deposit and can be run by any reader. MEDICAL ADVISORY: research prototype only — NOT a licensed medical device, NOT medical advice, and no drug is evaluated. Nobody should change any medication because of this work. No human evidence exists on benzodiazepine exposure and outcomes in diffuse midline glioma; the best available anti-seizure-medication evidence is null; no professional body has issued guidance. For medical emergencies in India: 112. Series: Part of the Maya-Vaidya Series — neuromorphic AI for clinical neuroscience. India-first. The Cancer sub-series arc: wire the tumour into the circuit (P7 Arbuda) → add the channel the circuit secretes (P8 Srava) → ask what sign the inhibitory channel carries (P9 Viparyaya). Links: GitHub Repository | Interactive Dashboard | FAQ | Full Series Index — venky2099.github.io Venkatesh Swaminathan · M.Sc. candidate, Data Science & AI, BITS Pilani · founder@nexuslearninglabs.inNexus Learning Labs, Bengaluru · UDYAM-KR-02-0122422 · BHASKAR IN-0526-9452JS · ORCID: 0000-0002-3315-7907 · VAIRAGYA_DECAY_RATE = 0.002315 — an ORCID-derived provenance mark, not an experimental parameter. Each paper's disclosure block states whether it reached a result in that paper. · Canary: MayaNexusVS2026NLL_Bengaluru