Disease evidence module · multidisease-evidence-v0.2.0

SMA · Proximal spinal muscular atrophy

Name: SMA evidence module
Creator: NMD-VCell

Human neural response evidence is separated from residual treated-muscle signals and a registered cell-resolution extension.

Measured neural response · line-aggregated extension

Open module API Compare diseases

Unified v0.2 contractstrict state + qualifiers · stable candidate IDs · unit structure · provenance · next experimentJSON Schema

Spinal organoids31 samples · 3 control + 2 SMA lines

Global ASO expression reversalNot observed · ρ=+0.10

Treated muscle8 SMA II + 7 controls

KIF5A after SMN loss−1.31 / −1.32 log2FC

Registered scRNA extension37,157 cells · 4 biological lines · directional

Interpretation and boundary

Positive, null and missing evidence remain separate.

GSE290980 adds 37,157 QC-retained cells after collapsing eight technical libraries to four biological lines. The largest marker-defined composition shift involved Motor_neuron_like. Effects remain directional (two lines per genotype), marker-defined identities require confirmation and no candidate is promoted to VALIDATED.

Selected evidence figure

Measured context and claim ceiling stay visible together.

SMA measured evidence and explicit claim boundary.

Analysis-plan-frozen single-cell extension · analyzed

GSE290980 localizes the signal without treating cells as replicates.

Summary JSON

Libraries8

Biological lines4

QC-retained cells37,157

Bounded localizationMotor_neuron_like composition shift

Source study finding

SMA organoids showed broad neurodevelopmental dysregulation, altered electrical activity and early SMN-directed ASO rescue in the source study.

Faravelli et al., Nature Communications (2026) · DOI 10.1038/s41467-025-67725-1 ↗

SMA analysis-plan-frozen single-cell extension with cell-state visualization and biological-line-level effects.

NMD-VCell reanalysis: technical-library aggregation followed by line-level directional effects without calibrated p-values. This extension refines state localization and the next experiment; it is not a public timestamped registration and does not validate a therapeutic target. The QC-table serialization repair is separately audited and did not modify statistical fields. The directly hosted disease vector is a byte-preserving gzip of the source SVG; raw vectors and standalone per-disease/combined PDFs remain checksum-addressed in the source release, while the synchronized v08.7 packet carries the publication figures.

Disease vector source (SVG.gz) Source release only · open manifestSynchronized manuscript/figure manifestQC audit PDF Source release only · open manifestAnalysis report Run-attempt audit QC serialization audit Analysis code

Dataset registry

Each object declares its biological/statistical unit.

MEASUREDGSE290979 ↗

INTERVENTION_RESPONSE

Human SMA type I spinal-cord organoids

31 bulk RNA-seq samples; 3 control and 2 SMA donor-derived lines

Unit: donor-derived organoid line; technical samples are not biological replicates
Role: Disease and R6-MO / scramble response at line level

MEASUREDGSE290980 ↗

SINGLE_CELL_REANALYSISLINE_LEVEL_DIRECTIONALN2_LINES_PER_GENOTYPE

Matched spinal/cerebral organoid single-cell data

8 single-cell libraries; registered as cell-resolution companion

Unit: donor-derived organoid line after technical-replicate aggregation; two lines per genotype
Role: Neural cell-state reference

MEASUREDGSE252128 ↗

DIRECTIONAL

Treated type II SMA paravertebral muscle

8 treated SMA vs 7 controls

Unit: independent skeletal-muscle biopsy
Role: Residual OXPHOS, denervation and fibrosis hypotheses; all module q>0.30 here

PERTURBEDGSE302774 ↗

Human motor and cortical neurons after SMN knockdown

3 replicates per condition in each neural context

Unit: independent culture replicate within each neuronal context
Role: Orthogonal KIF5A downstream mechanism evidence

Candidate establishment

Role-typed objects link directly to the next missing experiment.

Object	Role	State + qualifiers	Permitted interpretation / next experiment
SMN1 `NMDVCELL:SMA:CANDIDATE:SMN1` Identifiers & provenance HGNC:11117 · ENSG00000172062 Datasets: GSE290979, GSE252128, GSE302774, GSE290980 Unit structure: donor-derived organoid line; technical samples are not biological replicates; independent skeletal-muscle biopsy; independent culture replicate within each neuronal context; donor-derived organoid line after technical-replicate aggregation; two lines per genotype	etiologic_anchor	MEASURED SINGLE_CELL_STATE_LOCALIZATIONNO_TARGET_PROMOTION	Disease anchor Next missing experiment Retain as the disease reference axis; do not reinterpret it as a ranked therapeutic output.
SMN2 `NMDVCELL:SMA:CANDIDATE:SMN2` Identifiers & provenance HGNC:11118 · ENSG00000205571 Datasets: GSE290979, GSE290980 Unit structure: donor-derived organoid line; technical samples are not biological replicates; donor-derived organoid line after technical-replicate aggregation; two lines per genotype	validated_treatment_axis	CORRECTED SOURCE_REPORTEDSINGLE_CELL_STATE_LOCALIZATIONNO_TARGET_PROMOTION	Positive-control treatment axis; bulk expression did not globally reverse Next missing experiment Use as a positive-control axis while measuring residual cell-state and functional endpoints.
KIF5A `NMDVCELL:SMA:CANDIDATE:KIF5A` Identifiers & provenance HGNC:6323 · ENSG00000155980 Datasets: GSE302774, GSE290980 Unit structure: independent culture replicate within each neuronal context; donor-derived organoid line after technical-replicate aggregation; two lines per genotype	downstream_candidate	PERTURBED SINGLE_CELL_STATE_LOCALIZATIONNO_TARGET_PROMOTION	Mechanistic candidate; source-reported functional rescue, no patient efficacy Next missing experiment Perturb in line-aware motor-neuron models and test rescue plus toxicity endpoints.
OXPHOS_MODULE `NMDVCELL:SMA:CANDIDATE:OXPHOS_MODULE` Identifiers & provenance module/endpoint Datasets: GSE252128, GSE290980 Unit structure: independent skeletal-muscle biopsy; donor-derived organoid line after technical-replicate aggregation; two lines per genotype	residual_program	MEASURED DIRECTIONALSINGLE_CELL_STATE_LOCALIZATIONNO_TARGET_PROMOTION	Combination-therapy hypothesis space; no single target rank or significant module claim Next missing experiment Replicate the program in an independent treated cohort before resolving individual targets.

Supports

human neural disease/intervention response
source-reported SMN2 splicing and functional rescue
KIF5A downstream mechanism
residual-muscle hypotheses

Does not support

global gene-expression reversal
significant OXPHOS/denervation/fibrosis module claim in this reanalysis
patient motor-outcome prediction
combination-therapy ranking

Machine-readable module

Summary, datasets, candidate logic and claim boundary are separate objects.

summary · datasets · candidates · boundary