Spinal muscular atrophy · research datasets

Open SMA Research Data

Movement, electrophysiology, biomarker, RNA-sequencing, registry, and digital-measure resources described by cohort, available files, access level, and source.

Research use. These datasets represent specific cohorts and experimental settings. Consult each dataset card for licensing, study design, citation, and limitations.

Open · CC BY 4.0

Upper-limb Kinect movement

Seated reaching-game recordings collected across four visits from 18 ambulant adults with SMA type III and 19 controls.

  • Approximately 16 Hz 3D coordinates for nine upper-body joints
  • Task-event logs and derived movement features
  • Four PLOS supplements and four benchmark tables

37 participants · longitudinal movement data

View files and documentation →
Open · CC BY 4.0

Epidural spinal stimulation

Figure-source data from a first-in-human study of epidural spinal-cord stimulation in three adults with SMA type III.

  • Walking and gait measurements
  • Electromyography, strength, and stimulation-response data
  • CSV and spreadsheet files for Figures 2–6 and Extended Data Figures 2–10

3 participants · 28.2 MB archive

View files and documentation →
Open · CC BY 4.0

BforSMA biomarkers

Documents and result tables from a study of candidate proteins, metabolites, and transcripts in children with SMA and controls.

  • Five supplementary tables
  • Two appendices and one reported biomarker list
  • No participant-by-feature measurement matrix

108 children with SMA + 22 controls

View files and documentation →
Open · NCBI terms

Motor-neuron RNA sequencing

Processed RNA-sequencing results from human iPSC-derived motor neurons representing two SMA and two control cell lines.

  • Eight libraries: two replicates per cell line
  • Differential-expression and alternative-splicing outputs
  • GEO family metadata and SRA run inventory; raw reads are not included

GEO GSE108094 · SRA SRP126773

View files and documentation →
Application required

TREAT-NMD SMA registry

An international clinical-registry resource using a harmonized 120-item SMA core dataset.

  • Clinical and longitudinal registry variables
  • Public core-dataset definition
  • Participant data require a reviewed research enquiry

International registry network

View schema and access process →
Application required

Cure SMA Clinical Data Registry

A patient registry covering SMA natural history, treatments, clinical characteristics, and outcomes.

  • More than 1,000 reported participants
  • Longitudinal clinical and treatment information
  • Participant data use follows a governed research process

Clinical registry · application required

View registry information →
Sponsor controlled

JEWELFISH digital measures

Digital measurements collected in the JEWELFISH study to characterize motor and respiratory function.

  • Touchscreen tasks
  • Accelerometer, gyroscope, respiratory, and upper-limb signals
  • Participant-level streams are not publicly released

116 reported participants · sponsor controlled

View study description →
Upper-limb dataset

Reach-intent classification benchmark

Each completed trial corresponds to one of ten target positions. A model receives the first 25%, 50%, or 100% of the 3D skeleton sequence and predicts the target label from 0 to 9.

Participant splitFive fixed folds keep every participant entirely within train or test.
CalibrationOne-shot and five-shot settings use the first completed example of each target from a test participant.
Across visitsThe cross-visit setting calibrates on visit 0 and evaluates later visits.

Reference result: on the paper-QC subset, the five-fold nearest-centroid baseline reaches 84.08% accuracy at 25% movement progress across 3,655 trials. The dataset includes complete results for all checkpoints, calibration settings, macro recall, and SMA/control evaluation slices.

Scope

Study populations and limitations