← Back to home
GLA × Genetics (UK Biobank mGWAS)

Polygenic Control-Layer Erosion in ME/CFS

Interpreting signals from the UK Biobank metabolite genome-wide association study within the Gut–Liver–Autonomic (GLA) framework (L1–L5), with explicit separation of genetically constrained system architecture and epigenetically plastic control states.

Version: 2.4 Last updated: January 5th 2026 Author: Michael Daniels
one-sentence takeaway:

Patterns observed in the UK Biobank ME/CFS metabolite mGWAS are best interpreted as polygenic constraints that narrow physiological control headroom across multiple GLA layers, increasing vulnerability to epigenetically driven persistence and impaired recovery rather than indicating single-gene causation.

Section 1

Why this page exists

This document interprets genes highlighted by the UK Biobank ME/CFS metabolite genome-wide association study within the Gut–Liver–Autonomic (GLA) framework, without falling into “single gene → single mechanism” errors. The goal is a phenotype-safe interpretation: genetic signals are treated as modifiers of control headroom that shape which epigenetic and physiological states become stable under stress.

Boundary discipline (important):
  • mGWAS ≠ deterministic cause. It identifies statistical enrichment patterns that can alter vulnerability and recovery limits, not disease-specific mutations.
  • GLA layers are control layers. Genes are mapped to where they are most likely to influence control bandwidth, buffering capacity, routing, or recovery.
  • Symptoms emerge downstream. Interpretation proceeds as gene → dominant control layer → failure mode → symptom domain, rather than gene-to-symptom causation.
↑ Back to top
Section 2

Canonical claim

Genes highlighted by the UK Biobank ME/CFS metabolite mGWAS are interpreted here as sources of polygenic constraint that reduce stability margins across multiple GLA control layers. These constraints do not create illness in isolation, but they shape how physiological systems respond to infection, stress, and recovery demands.

  • Easier to enter persistent immune, autonomic, metabolic, or vascular control states following an initiating stressor.
  • Harder to exit those states once ER stress, redox imbalance, or NAD⁺-dependent recovery ecology is degraded.
  • More clinically heterogeneous, because small differences in which control layer is most constrained shift the dominant symptom expression and crash pattern.

This framing provides the bridge between polygenic enrichment signals and the GLA concept of baseline threshold erosion, without implying genetic determinism or single-pathway causation.

↑ Back to top
Section 3

Study and method framing: UK Biobank metabolite GWAS in ME/CFS

This section defines how metabolite genome-wide association signals are interpreted within the Gut–Liver–Autonomic (GLA) framework, and clarifies the methodological scope and limits of the UK Biobank ME/CFS mGWAS used in this analysis.

Study summary

We draw on the metabolite genome-wide association study (mGWAS) conducted by Huang et al. in the UK Biobank to examine genetic contributions to metabolic heterogeneity in ME/CFS. The study tested associations between genetic variants and circulating metabolic biomarkers in individuals with ME/CFS versus healthy controls, rather than testing direct associations with disease status.

The analysis evaluated 135 serum metabolic traits measured by high-throughput nuclear magnetic resonance spectroscopy in 875 individuals with ME/CFS and 36,033 controls, identifying SNP–metabolite associations within each cohort. Findings are interpreted here as distributed, state-dependent regulatory modulation rather than evidence of single-gene causation.

What “mGWAS signals” mean in this document

  • mGWAS signals are treated as probabilistic modifiers of regulatory pathways (control headroom and stability), not as binary disease switches or causal mutations.
  • Gene interpretation proceeds in two steps: (1) gene → primary physiological function, followed by (2) function → dominant GLA layer.
  • Given the marked female predominance of ME/CFS in the UK Biobank cohort, interpretations must remain compatible with sex-biased regulatory risk, including differential susceptibility to membrane instability and control-layer erosion.

Study scope and interpretive boundaries

The metabolite GWAS conducted by Huang et al. analyzed genetic associations with 135 circulating metabolic biomarkers measured by high-throughput nuclear magnetic resonance spectroscopy in individuals with ME/CFS and healthy controls drawn from the UK Biobank. The study was designed to investigate genetic contributions to metabolic variation, rather than to identify disease-causing variants.

Identified associations exhibited small effect sizes and substantial overlap with regulatory genes observed in control populations, often differing in magnitude or direction between cohorts. These features indicate state-dependent genetic modulation rather than disease-specific causation.

Within the GLA framework, such findings are interpreted as evidence of polygenic control-layer erosion, in which distributed genetic variation alters the fidelity, timing, buffering, and recovery properties of physiological systems.

Primary study reference

Huang K, Muneeb M, Thomas N, Schneider-Futschik EK, Gooley PR, Ascher DB, Armstrong CW. Exploring a genetic basis for the metabolic perturbations in ME/CFS using UK Biobank. iScience, 2025.
https://www.sciencedirect.com/science/article/pii/S2589004225025775

Genes highlighted in the UK Biobank ME/CFS mGWAS

The following genes were identified through significant SNP–metabolite associations and form the basis of subsequent GLA layer mapping:

NLRC5 · ADAP1 · TMEM258 · HERPUD1 · MYRF · NR1H3 (LXRα) · CETP · LIPC · LIPG · AQP9 · AQP7 · APOE · LPL · SCGN · SLC18A1 (VMAT1) · DDR1 · CPS1

Pathway and functional enrichment themes reported

  • Regulation of lipid transport and lipoprotein remodeling
  • Metabolic routing and substrate redistribution rather than ATP generation
  • Neuroendocrine and monoaminergic handling
  • Vascular buffering, perfusion stability, and recovery-related processes

Notably, classical inflammatory cytokine pathways and core mitochondrial energy production pathways were not preferentially enriched, supporting a regulatory control interpretation.

↑ Back to top
Figure 1 — Canonical Gene → Layer → Symptom Map (GLA L1–L5)

Single-map schematic of how Armstrong-style mGWAS signals are interpreted as control-layer erosion modifiers within a layered physiology model (L1–L5). This figure encodes the fixed mapping discipline: dominant function → dominant layer, with downstream symptom domains shown as emergent expressions rather than single-pathway causation.

How to read this figure
  • Middle column: each layer (L1–L5) is defined by its locked role in control, membranes, routing, buffering, or recovery.
  • Mapped genes: each gene is assigned to one dominant layer by primary physiological function (not by symptoms).
  • Right column: “emergent domains” are clinical expressions that arise downstream from layered control erosion — not one-gene causes.
GLA mapping discipline (locked L1–L5 interpretations) From mGWAS variants → Layer role → Emergent symptom domains Signals are treated as modifiers of control / routing / buffering / recovery, not single-pathway causes. Control-layer erosion ≠ immune activation. Layers constrain interpretation: dominant function → dominant layer; cross-layer effects modeled downstream. GLA Layers (L1–L5) L1 Immune Control / Signal Termination Locked role: duration / routing / termination fidelity. Interpretation: control-state instability & impaired exit, not cytokine excess. Mapped genes NLRC5 • ADAP1 • TMEM258 • HERPUD1 L2 Membrane / Execution Surface Stability Locked role: baseline execution-surface resilience. Interpretation: slow, cumulative, recovery-limiting (not hyper-excitability). Mapped gene MYRF L3 BA-GLA Routing / Substrate Throughput Locked role: redistribution & timing of lipid/substrate flux. Interpretation: normal fasting labs can coexist with post-load instability. Mapped genes NR1H3 (LXRα) • CETP • LIPC/LIPG • AQP9 • AQP7 L4 Distribution / Buffering / Execution Coupling Locked role: systemic spread, buffering, autonomic/hormonal coupling. Interpretation: severity & timing modulation; not an initiating layer. Mapped genes APOE • LPL • SCGN • SLC18A1 (VMAT1) L5 Perfusion Stability / Recovery Bandwidth Locked role: recovery depth, reset capacity, perfusion stability. Interpretation: delayed PEM & cumulative intolerance reflect recovery failure. Mapped genes DDR1 • CPS1 Emergent domains (expressions, not one-gene causes) Signal-duration instability Prolonged priming Poor termination May occur without high circulating cytokines Execution-surface fragility Reduced membrane repair headroom Routing / timing mismatch Post-load lipid and substrate redistribution issues Normal fasting labs can persist Systemic spread & coupling Autonomic volatility Multi-system crash timing Severity modulator Recovery failure bandwidth Delayed PEM Cumulative intolerance Perfusion / reset stability limits Canonical takeaway encoded by this figure mGWAS signals map preferentially to genes that regulate control quality (L1), execution-surface stability (L2), routing/throughput timing (L3), systemic buffering & coupling (L4), and recovery bandwidth (L5), supporting a layered control-erosion interpretation rather than a single immune-inflammatory cause.

Figure 1. Canonical gene → layer → symptom mapping for the GLA framework (L1–L5). Genes are assigned to one dominant layer by a fixed mapping discipline (dominant function, timescale, failure mode, single-layer constraint). Symptom domains on the right represent emergent expressions of layered control erosion and downstream buffering/recovery limits, not one-gene causes. This figure is intended to be referenced directly by the Methods logic and reused as the schematic anchor for Results and Discussion.

↑ Back to top
Section 5

Evidence plane: compact gene → layer mapping table

This table is the “Methods in miniature.” Each gene is assigned to one primary layer by fixed rules (dominant function, timescale, failure mode). “Emergent domains” describe downstream expressions (non-reductionist), not single-gene causes.

Gene Primary layer Control role (locked) Emergent domains (non-reductionist)
NLRC5 L1 Immune set-point / duration control; couples immune state to lipid–membrane traits Signal-duration instability; prolonged priming windows; downstream layer spillover likelihood
ADAP1 L1 Signal routing / spillover amplifier from immune control into execution layers Higher propagation of flares into membrane/buffering domains; variability in crash initiation thresholds
TMEM258 L1 ER-glycosylation / signal-quality regulator (receptor folding/half-life; noisy or prolonged signaling risk) Noisy immune signaling; longer primed states; downstream threshold lowering
HERPUD1 L1 Stress-resolution / recovery-brake gene (ERAD; signal termination support) Persistence tendency (“stuck-on” states); baseline erosion pressure via poor resolution
MYRF L2 Structural membrane maintenance (sphingomyelin; long-term lipid-raft integrity) Execution-surface fragility; reduced repair headroom; cumulative intolerance
NR1H3 (LXRα) L3 Transcriptional routing/buffering node linking immune tone to hepatic lipid handling Post-load redistribution vulnerability; delayed symptom amplification via routing constraints
CETP L3 Lipid redistribution modifier across lipoprotein compartments Vascular/membrane buffering load-shifts; delayed PEM susceptibility via redistribution timing
LIPC / LIPG L3 Throughput-conditioning lipases remodeling lipoproteins & endothelial lipid exposure Timing-dependent membrane repair substrate availability; vascular tolerance variability
AQP9 L3 Substrate-flux gain channel (glycerol/lactate/urea movement; accelerates load) Delayed PEM amplification when buffering is limited; load acceleration effects
AQP7 L3 Substrate-availability timing channel (glycerol release; baseline access & recovery depth) Recovery depth modulation; variability in baseline energy access after load
APOE L4 Vascular lipid buffer-size & clearance modifier (cumulative tolerance) Delayed PEM magnitude via buffering; systemic spread vs containment tendency
LPL L4 Capillary fuel-access gatekeeper (tissue-level energy availability) Early fatigue / ischemic metabolism bias when constrained; crash style variability
SCGN L4 Ca²⁺-dependent execution-gain amplifier at neuroendocrine interface (synchrony/overshoot risk) Autonomic volatility; multi-system synchrony; overshoot susceptibility under fragile control
SLC18A1 (VMAT1) L4 Execution precision / containment regulator (monoamine packaging fidelity) Crash variability; autonomic volatility without changing neurotransmitter synthesis
DDR1 L5 Vascular mechanosensing / stiffness interpretation (posture/load-dependent stress integration) Orthostatic intolerance tendency; perfusion instability under load
CPS1 L5 Metabolic recovery bandwidth bottleneck (urea-cycle nitrogen handling; recovery depth limits) Delayed PEM depth; incomplete reset; cumulative intolerance under repeated stressors
↑ Back to top
Section 6

Detailed layer summaries (L1 → L5)

Each gene is assigned to one primary GLA layer by fixed rules (dominant function, timescale, and failure mode). Layer summaries below are written to prevent drift: they define what the layer does, how it fails, which genes map there, and what symptom domains can emerge downstream (non-reductionist).

L1 — Immune control layer (signal quality & termination)

Core function: Regulates signal duration, routing, and termination fidelity of innate immune activation (i.e., whether activation resolves as a pulse or persists as a control state).

failure mode: impaired exit from activation (signal-duration failure / prolonged priming) rather than cytokine excess.

Mapped genes (primary): NLRC5 · ADAP1 · TMEM258 · HERPUD1

  • Downstream expression (emergent): prolonged “primed windows” and overlapping recovery periods, even without high circulating cytokines.
  • System consequence: increased probability that modest stressors propagate into downstream layers (membrane, routing, buffering).
  • Clinical domains shaped: flare frequency, persistence tendency, and variability in “on/off” control-state transitions.

L2 — Membrane / execution-surface stability (baseline resilience)

Core function: Maintains execution-surface integrity (lipid-raft and membrane organization) and long-term resilience of signaling interfaces.

failure mode: baseline execution-surface fragility that is slow, cumulative, and recovery-limiting (not hyper-excitability).

Mapped gene (primary): MYRF

  • Downstream expression (emergent): reduced tolerance to repeated activation because membrane repair / re-stabilization headroom is limited.
  • System consequence: downstream layers (routing, buffering, recovery) must absorb more stress because the interface itself is less stable.
  • Clinical domains shaped: cumulative intolerance, delayed recovery after episodes, and “baseline erosion” pressure under repeated load.

L3 — BA-GLA routing / substrate throughput (redistribution & timing)

Core function: Controls redistribution and timing of lipid/substrate handling (routing and throughput), shaping how stress is absorbed and reallocated across compartments rather than “produced.”

failure mode: routing / timing mismatch (state-dependent) that amplifies post-load instability despite normal fasting labs.

Mapped genes (primary): NR1H3 (LXRα) · CETP · LIPC / LIPG · AQP9 · AQP7

  • Downstream expression (emergent): delayed or disproportionate post-exertional shifts in lipid/substrate availability and vascular exposure.
  • System consequence: stress becomes redistributed in ways that increase downstream buffering burden (L4) and recovery burden (L5).
  • Clinical domains shaped: delayed PEM timing, “normal baseline labs” paradox, and variability in cognitive/vascular vs metabolic symptom dominance.

L4 — Distribution, buffering & execution coupling (systemic expression)

Core function: Determines how upstream signals and substrates are expressed systemically via buffering capacity, capillary fuel access, and neuroendocrine/autonomic coupling.

failure mode: systemic spread and timing amplification—this layer modulates severity and synchrony; it does not initiate pathology.

Mapped genes (primary): APOE · LPL · SCGN · SLC18A1 (VMAT1)

  • Downstream expression (emergent): multi-system crash patterns and autonomic volatility from tighter coupling between signals and execution outputs.
  • System consequence: buffering limits determine whether stress remains localized or becomes systemic (spread/containment problem).
  • Clinical domains shaped: orthostatic/autonomic instability, crash “style” variability, and delayed symptom amplification magnitude.

L5 — Perfusion stability & recovery bandwidth (reset capacity)

Core function: Governs recovery depth (reset capacity), perfusion stability integration, and metabolic recovery bandwidth after load.

failure mode: recovery failure (impaired reset) rather than excessive execution—biasing toward delayed PEM and cumulative intolerance.

Mapped genes (primary): DDR1 · CPS1

  • Downstream expression (emergent): delayed PEM severity reflects inability to regain baseline after load (recovery bandwidth bottleneck).
  • System consequence: posture/load-dependent perfusion stress and incomplete metabolic reset accelerate baseline threshold erosion over time.
  • Clinical domains shaped: orthostatic intolerance persistence, delayed crash depth, and phase progression under repeated incompletely buffered events.

Downstream outcome (not a layer): Post-Exertional Malaise (PEM)

PEM is treated as an integrated downstream expression of constraints across L1–L5: signal-duration instability (L1) + execution-surface fragility (L2) + routing/timing mismatch (L3) + systemic buffering/coupling (L4) + recovery bandwidth limits (L5).

  • Why it’s not a layer: PEM describes the system-level output when layered control and recovery limits are exceeded.
  • Why timing is delayed: redistribution (L3), buffering/coupling (L4), and reset limits (L5) can amplify symptoms after the initiating load has passed.
  • Why patterns differ: heterogeneity reflects which layer becomes rate-limiting in a given patient and phase, not different diseases.
↑ Back to top
Section 7

Phenotype lens: SMPDL3B-shedding vs SMPDL3B-deficient (and why sex ratio matters)

SMPDL3B-Shedding (women-biased risk)

  • Failure mode: threshold lowering → episodic defensive overshoot (PI-PLC shedding episodes).
  • Control signature: oscillatory instability; recovery possible if overshoot pressure drops.
  • Interpretation for mGWAS: signals that lower execution thresholds or prolong primed windows bias toward episodic shedding.

SMPDL3B-Deficient

  • Failure mode: attractor shift → chronically low anchoring baseline; low recovery bandwidth.
  • Control signature: capacity drains under amplification; difficulty returning to baseline.
  • Interpretation for mGWAS: signals that reduce recovery ecology (NAD⁺/ER/REDOX) or anchoring stability bias toward deficiency-like control collapse.
Sex ratio constraint: If ~75% of the cohort is women, the combined signal set must be compatible with a population that is more likely to express shedding-prone oscillation early (hormone-biased gain), even if late severe states can converge clinically.

Gene-level bias toward shedding vs deficiency failure modes

Genes highlighted in the UK Biobank ME/CFS mGWAS can be grouped by the type of control failure mode they would bias toward if regulatory stability is lost. These groupings reflect dominant physiological roles and failure signatures, not deterministic phenotype assignment.

Shedding-bias (threshold-lowering / overshoot-prone): NLRC5 · ADAP1 · TMEM258 · SCGN · SLC18A1 (VMAT1)

Deficiency-bias (recovery-limiting / baseline erosion): MYRF · HERPUD1 · NR1H3 (LXRα) · CETP · LIPC/LIPG · AQP9 · AQP7 · APOE · LPL · DDR1 · CPS1

These biases describe how control failure is likely to express under stress (episodic overshoot versus chronic low anchoring), not which phenotype a given individual must exhibit.

Sex-stratified findings in the mGWAS

The UK Biobank ME/CFS mGWAS did not report sex-stratified association analyses or gene-by-sex interaction testing. No gene was identified as having a female-specific or male-specific genetic effect.

Accordingly, sex-related interpretation in this document reflects established biological constraints (e.g., immune gain and SMPDL3B shedding dynamics) applied to a female-predominant cohort, rather than sex-specific genetic claims derived from the mGWAS itself.

Why a female-predominant cohort appears shedding-weighted

Independent of genetic association testing, established biological differences in immune and neuroendocrine regulation help explain why a female-predominant ME/CFS cohort is more likely to express shedding-compatible instability early in disease.

  • Women, on average, exhibit higher immune and autonomic gain, increasing susceptibility to threshold-lowering and episodic control overshoot rather than silent capacity exhaustion.
  • These gain differences favor oscillatory, stimulus-sensitive crashes that are more visible, reportable, and likely to be captured in large population datasets.
  • Over time, repeated overshoot without full recovery can erode baseline stability, allowing late severe disease to converge clinically with deficiency-pattern states without implying identical underlying mechanisms.

This context explains cohort weighting and signal visibility; it does not imply sex-specific genetic causation or exclude SMPDL3B-deficient mechanisms in men or women.

↑ Back to top
Section 8

Genetic constraint vs epigenetic plasticity across layers

This section formalizes the distinction between genetic constraint and epigenetic plasticity within the GLA framework. Genetic variation defines the structural and regulatory boundaries of physiological systems, while epigenetic and state-dependent factors determine where the system operates within those boundaries over time, including during phase progression.

  • More genetically constrained (architecture): baseline pathway capacity, membrane and lipid-raft organization, receptor and signaling set-points, buffering limits.
  • More epigenetically plastic (state): immune priming duration, ER/ROS/NAD⁺ recovery ecology, autonomic lock-in, symptom dominance shifts across phases.

Canonical plasticity map across GLA layers

GLA layer Genetic constraint (sets boundaries) Epigenetic / state plasticity (moves within bounds)
L1 — Immune control Signal gain tendencies, termination efficiency, immune set-point architecture Duration of priming, phosphatase inhibition, persistence vs resolution of activation
L2 — Membrane stability Structural lipid-raft integrity, anchoring and repair capacity Accumulated membrane stress, recovery completeness between episodes
L3 — Routing & throughput Lipid and substrate redistribution limits, buffering architecture Timing mismatches, post-load redistribution, delayed metabolic instability
L4 — Distribution & buffering Vascular buffering size, neuroendocrine coupling gain Autonomic volatility, crash spread vs containment, symptom synchronization
L5 — Recovery bandwidth Maximum recovery depth, perfusion integration limits Reset completeness, delayed PEM severity, baseline erosion over repeated stress

Implications for heterogeneity and disease progression

Within this framework, genetic variants do not determine disease state directly. Instead, they shape the control landscape within which epigenetic stressors, environmental exposures, and recovery failures operate.

Phase progression reflects cumulative movement toward the limits of these genetically constrained boundaries, driven by incomplete recovery rather than new pathology. Conversely, symptom improvement reflects restoration of control within existing constraints, not reversal of genetic architecture.

↑ Back to top
Section 9

Predictions and measurements

This section outlines testable predictions that follow directly from the GLA gene-to-layer mapping. Because genetic variants are fixed while disease expression evolves, the framework predicts layer-linked heterogeneity, state dependence over time, and phenotype-consistent patterns that can be evaluated using longitudinal and stress-responsive measurements.

If the mapping is correct, we should observe

  • Layer-linked heterogeneity: dominant symptom clusters should align with the most genetically constrained layer (e.g., immune control, membrane stability, routing, buffering, or recovery), rather than with a single disease subtype.
  • State dependence over time: worsening disease phase should reflect increasing overlap of primed windows and reduced recovery completeness, even though underlying genetic architecture remains unchanged.
  • Phenotype-consistent patterns: SMPDL3B-shedding should manifest as oscillatory, stimulus-sensitive episodes, whereas SMPDL3B-deficient states should show persistent baseline instability and low recovery bandwidth.

Layer-specific measurement expectations

GLA layer Representative measurements Expected direction with worsening control
L1 — Immune control Interferon-stimulated gene expression dynamics, pSTAT duration, immune recovery kinetics Prolonged signal duration, incomplete termination, increased primed-window overlap
L2 — Membrane stability Lipid raft organization proxies, membrane repair markers, anchoring stability Reduced resilience to repeated load, slower or incomplete re-stabilization
L3 — Routing & throughput Lipoprotein subclass shifts, bile-acid and lipid redistribution profiles Increasing post-exertional mismatch despite near-normal fasting values
L4 — Distribution & buffering Vascular tolerance measures, autonomic variability, neuroendocrine coupling indices Greater systemic spread of stress, increased crash synchrony and severity
L5 — Recovery bandwidth Recovery time constants, perfusion stability, delayed PEM depth Slower and shallower recovery, cumulative baseline erosion over time

Phase-linked predictions

  • Early phase: genetic constraints shape which layer fails first, but recovery between episodes remains largely intact.
  • Intermediate phase: repeated incomplete recovery produces overlap of primed windows, amplifying post-exertional instability without new pathology.
  • Late phase: recovery bandwidth becomes the dominant limiter, leading to persistent baseline instability and convergence of clinical severity across phenotypes.

These phase transitions are predicted to occur without changes in genetic architecture, reflecting progressive movement toward genetically constrained control boundaries rather than the emergence of new disease mechanisms.

↑ Back to top
Figure X — From polygenic modifiers to phase drift: gene → layer → measurement

Genes bias layer constraints; measurements report state; phase progression reflects drift within fixed genetic boundaries.

POLYGENIC MODIFIERS Genes (mGWAS) Small-effect variants Not disease-causing switches NLRC5 ADAP1 TMEM258 HERPUD1 MYRF NR1H3 CETP LIPC/LIPG AQP9 AQP7 APOE LPL SCGN SLC18A1 DDR1 CPS1 Genes define constraints → GLA ARCHITECTURE Dominant layer constraints (L1–L5) One primary layer per gene L1 — Immune control Signal duration & termination L2 — Membrane stability Execution surface integrity L3 — Routing & throughput Substrate redistribution timing L4 — Buffering & coupling Spread vs containment L5 — Recovery bandwidth Reset depth & perfusion stability Layers → measurable state → MEASURABLE SIGNALS Measurements (state, not cause) Stress-responsive & longitudinal L1 measures pSTAT duration · ISG kinetics termination / primed-window overlap L2 measures raft stability proxies · re-anchoring L3 measures lipoprotein subclass shifts BA / lipid redistribution timing L4 measures autonomic variability · vascular tolerance L5 measures recovery time constants · delayed PEM depth State trajectories → phase drift PHASE DRIFT Same genes, different state P1 P2 P3 P4 Genetic boundary Phase drift over time Phase 1 wide recovery Phase 2 partial recovery Phase 3 baseline erosion Phase 4 recovery failure Genes fixed State drifts toward boundary Architecture / constraint Measured state (indirect) State drift over phases
Figure X. Polygenic modifiers identified by metabolite GWAS map to dominant GLA layers (L1–L5), which define constraint boundaries. Layer-linked measurements report dynamic state within these constraints. Phase progression reflects drift toward constraint limits without changes in genetic architecture.
Section 10

References and interpretive context

Primary study referenced

Huang K, Muneeb M, Thomas N, Schneider-Futschik EK, Gooley PR, Ascher DB, Armstrong CW. Exploring a genetic basis for the metabolic perturbations in ME/CFS using UK Biobank. iScience, 2025.
https://www.sciencedirect.com/science/article/pii/S2589004225025775

This metabolite genome-wide association study (mGWAS) provides the genetic signal set interpreted in the present analysis. No additional external datasets were used.

Interpretive framework documents (GLA v2.1 → v2.4)

The following documents define the conceptual and methodological framework used to interpret genetic signals in this paper. They establish layer boundaries, phenotype discipline, and phase dependence, and are provided for transparency rather than as evidentiary sources.

These documents are intended to clarify interpretation boundaries and prevent over-extension of genetic associations into claims of causation, uniform pathology, or phase-invariant mechanisms.

↑ Back to top
Appendix A

Per-gene notes

Each block summarizes one highlighted gene using a fixed mapping discipline: gene → primary function → GLA layer assignment → phenotype bias → symptom route. These summaries describe probabilistic control effects, not deterministic causation.

NLRC5

  • Primary function: innate immune regulation; sets immune activation and termination setpoints.
  • GLA layer: L1 — immune control (signal gain and duration).
  • Phenotype bias: shedding-biased (threshold lowering / prolonged primed windows), particularly in early or oscillatory states.
  • Symptom route: immune priming → lipid/membrane-linked biomarker shifts → episodic post-exertional instability.

CD40

  • Primary function: immune co-stimulatory receptor; amplifies immune–immune and immune–vascular signaling.
  • GLA layer: L1 — immune control (signal amplification), with secondary influence on L5 buffering via immune–vascular coupling.
  • Phenotype bias: shedding-biased (overshoot-prone escalation under stress), state-dependent.
  • Symptom route: immune activation ↔ lipid transport changes ↔ vascular/perfusion instability during flares.

ADAP1

  • Primary function: immune signaling adaptor; modulates signal routing and execution efficiency at membranes.
  • GLA layer: L1 — immune control (routing and gain), bridging to L2 membrane execution.
  • Phenotype bias: shedding-biased (enhanced signal spillover rather than baseline failure).
  • Symptom route: immune activation → membrane execution spillover → oscillatory symptom amplification.

TMEM258

  • Primary function: ER-dependent protein glycosylation; determines immune receptor quality and stability.
  • GLA layer: L1 — immune control (signal quality and termination), interfacing with L2 membrane stability.
  • Phenotype bias: shedding-biased (noisy or prolonged signaling increases overshoot risk).
  • Symptom route: impaired immune signal fidelity → membrane instability → episodic crashes under stress.

HERPUD1

  • Primary function: ER-associated degradation and stress resolution; enables recovery from immune activation.
  • GLA layer: L1 — immune control (signal termination and recovery), interfacing with L2 membrane repair.
  • Phenotype bias: deficiency-biased (recovery-limiting, baseline erosion).
  • Symptom route: unresolved ER stress → prolonged immune activation → delayed recovery and cumulative PEM.

MYRF

  • Primary function: structural lipid and sphingomyelin program maintenance; long-timescale membrane integrity.
  • GLA layer: L2 — membrane and lipid-raft stability (execution surface architecture).
  • Phenotype bias: deficiency-biased (baseline fragility and limited re-anchoring capacity rather than overshoot).
  • Symptom route: reduced execution-surface resilience → poor tolerance to repeated stress → delayed recovery and cumulative PEM.

NR1H3 (LXRα)

  • Primary function: transcriptional coordination of lipid handling, bile-acid–linked routing, and immune–hepatic crosstalk.
  • GLA layer: L3 — BA-GLA metabolic routing and throughput capacity.
  • Phenotype bias: deficiency-biased (chronic under-support and constrained recovery bandwidth).
  • Symptom route: maladaptive lipid routing under immune stress → delayed metabolic clearance → post-exertional symptom amplification.

CETP

  • Primary function: redistribution of cholesteryl esters and triglycerides among lipoproteins (HDL ↔ LDL/VLDL).
  • GLA layer: L3 — BA-GLA lipid redistribution and routing.
  • Phenotype bias: deficiency-biased (buffering and repair substrate misallocation rather than gain).
  • Symptom route: post-stress lipid redistribution → endothelial and membrane stress → delayed cognitive and vascular symptoms.

LIPC / LIPG

  • Primary function: hepatic and endothelial lipoprotein remodeling; conditioning of particle quality and clearance timing.
  • GLA layer: L3 — BA-GLA throughput conditioning and lipid traffic processing.
  • Phenotype bias: deficiency-biased (baseline erosion via over-processing or mistimed clearance).
  • Symptom route: reduced phospholipid availability for repair → endothelial vulnerability → cumulative intolerance rather than sharp crashes.

AQP7

  • Primary function: glycerol release from adipose tissue; controls energy substrate availability.
  • GLA layer: L3 — BA-GLA substrate release and metabolic routing.
  • Phenotype bias: deficiency-biased (limited substrate availability and reduced recovery depth).
  • Symptom route: constrained energy mobilization → gradual fatigue accumulation → poor tolerance to sustained activity.

AQP9

  • Primary function: rapid glycerol, lactate, and solute flux in hepatocytes and immune cells.
  • GLA layer: L3 — BA-GLA substrate flux and routing speed.
  • Phenotype bias: deficiency-biased (throughput exceeding buffering capacity, especially in later phases).
  • Symptom route: accelerated metabolic influx → redox/NAD⁺ stress → delayed PEM following exertion or immune activation.

APOE

  • Primary function: lipid transport, clearance, and delivery to vascular, neural, and peripheral tissues.
  • GLA layer: L4 — vascular distribution and buffering capacity.
  • Phenotype bias: deficiency-biased (buffer-size limitation rather than execution gain).
  • Symptom route: reduced lipid buffering → endothelial and membrane vulnerability → delayed and cumulative PEM without overt dyslipidemia.

LPL

  • Primary function: hydrolysis of circulating triglycerides to enable tissue lipid-fuel uptake.
  • GLA layer: L4 — capillary fuel delivery and tissue energy access.
  • Phenotype bias: deficiency-biased (fuel access limitation rather than overshoot).
  • Symptom route: impaired lipid fuel delivery → early fatigue, ischemic metabolism, delayed PEM despite normal circulating lipids.

SCGN (Secretagogin)

  • Primary function: Ca²⁺-dependent neuroendocrine and autonomic secretion gain.
  • GLA layer: L4 — autonomic and neuroendocrine execution coupling.
  • Phenotype bias: shedding-biased (execution overshoot and oscillatory instability).
  • Symptom route: amplified stress-responsive secretion → multi-system crash patterns and episodic autonomic volatility.

SLC18A1 (VMAT1)

  • Primary function: vesicular monoamine packaging and controlled neurotransmitter/hormone release.
  • GLA layer: L4 — execution precision and autonomic containment.
  • Phenotype bias: shedding-biased (poor execution containment rather than baseline loss).
  • Symptom route: imprecise monoamine handling → autonomic volatility, stress intolerance, sensory and GI instability.

DDR1

  • Primary function: mechanosensing of extracellular matrix and vascular structural stress.
  • GLA layer: L5 — perfusion stability and mechanical stress interpretation.
  • Phenotype bias: shedding-biased (stress-signal amplification under load).
  • Symptom route: heightened response to posture or mechanical strain → orthostatic intolerance and load-dependent symptom flares.

CPS1

  • Primary function: mitochondrial urea-cycle throughput and nitrogen detoxification.
  • GLA layer: L5 — metabolic recovery bandwidth and post-exertional clearance.
  • Phenotype bias: deficiency-biased (recovery bottleneck rather than overshoot).
  • Symptom route: constrained nitrogen handling → prolonged mitochondrial stress, delayed PEM, poor recovery despite normal baseline labs.
↑ Back to top