About SMA Research Platform
The SMA Research Platform is an autonomous, multi-method computational discovery engine for Spinal Muscular Atrophy. We combine molecular screening, protein structure prediction, and literature-derived evidence analysis to nominate therapeutic candidates for experimental validation. All findings are computational predictions or hypotheses; none have been wet-lab validated.
Every data point on this platform comes from the live database — nothing is hardcoded. When new papers are ingested or compounds are screened, the entire platform updates automatically.
Methodology
The platform ingests literature daily from PubMed, bioRxiv/medRxiv preprints, ClinicalTrials.gov, and Google Patents. Each abstract passes a two-layer quality filter: first an SMA-relevance gate (must mention SMA, SMN, motor neuron, or approved therapy names), then a post-extraction quality gate that rejects claims about unrelated diseases.
Claims are extracted by AI (Claude/Gemini) in structured form: subject → predicate → object, with confidence score, evidence level, and tissue context. Each claim is linked back to its source paper. Targets are scored across 8 dimensions (expression evidence, clinical precedent, safety, novelty, convergence, druggability, disease relevance, literature volume) and ranked for drug discovery.
Computational biology pipeline: multi-method virtual screening (AutoDock Vina, DiffDock), AlphaFold3 and Boltz-2 protein–protein interaction modeling, ML proxy surrogate model (Random Forest), and molecular dynamics simulations for prioritized hits.
Data Sources
- PubMed — research literature (automated daily ingestion)
- ClinicalTrials.gov — clinical trial registry
- ChEMBL — compound bioactivity database
- STRING-DB — protein-protein interactions
- KEGG — biological pathways
- AlphaFold DB — predicted protein structures
- GPU fleet — DiffDock, MD simulations, mBER nanobody design
Technology
- Backend: FastAPI + PostgreSQL + asyncpg
- Frontend: Next.js + TypeScript + Tailwind CSS
- AI: Claude, Gemini, multi-LLM claim extraction
- Compute: GPU fleet — AF3, Boltz-2, Vina, MD simulations
- Pipeline: multi-stage with knowledge gates (BIO-GATE, DOCK-GATE, quality filters)
License
Open source under AGPL-3.0. Built by Christian Fischer / Bryzant Labs.
Citation
If you use data from this platform, please cite:
Fischer, C. (2026). SMA Research Platform — Open Evidence Graph for Spinal Muscular Atrophy. https://sma-research.info Bryzant Labs. Accessed [date].
BibTeX
@misc{fischer2026sma,
author = {Fischer, Christian},
title = {{SMA Research Platform --- Open Evidence Graph for SMA}},
year = {2026},
url = {https://sma-research.info},
note = {Accessed: 2026-04-11}
}