Manifold Causal Models Reveal Cryptic Drivers in Breast Cancer
Key findings: ASCL2 ranked #15 vs #15,000+, P=0.0048, validated in 6,253 patients.
Publications
A Curated High-Fidelity Carbide Materials Dataset (HF-CCD)
Manuscript: DIB-D-25-02507.
CSD-EWS: Pole & Damping Ratio Simulator
Explore σ, ω, and ζ on the s-plane with unified weak-damping (5%) alerts and bilingual UI.
regional-era5-chunks
Memory-efficient parallel bounding-box extraction for global ERA5 NetCDF datasets — avoids full-grid RAM overload when preparing atmospheric forcing for physical foundation models.
Datasets
TCGA Breast Cancer - Manifold Analysis
526 patients, 17,818 genes.
Format: CSV.
[Download link when published]
High-Fidelity Carbide Dataset
Curated structures with quality control.
[Download link when published]
Component Library
Total: 10 components | Format: JSON | License: MIT
Causal Drivers
ASCL2 — Transcription Factor
Evidence: STAR STAR STAR STAR STAR GOLD
Cancer type: Breast (Basal-like)
Function: Stemness activation, metastasis driver
[Download JSON]
OR10AG1 — Olfactory Receptor
Evidence: STAR STAR STAR STAR HIGH
Cancer type: Breast
Function: Survival regulation
[Download JSON]
NAPRT1 — Metabolic Enzyme
Evidence: STAR STAR STAR STAR HIGH
Cancer type: Breast
Function: NAD+ biosynthesis
[Download JSON]
Phenotypic Markers
ROPN1 — Basal-like Marker
Evidence: STAR STAR STAR STAR HIGH
Cancer type: Breast (TNBC)
Function: Phenotypic marker, 90% TNBC expression
[Download JSON]
FABP7 — Fatty Acid Binding Protein
Evidence: STAR STAR STAR STAR HIGH
Cancer type: Breast (metastasis)
Function: Brain metastasis marker
[Download JSON]
HORMAD1 — Meiosis Protein
Evidence: STAR STAR STAR MEDIUM
Cancer type: Breast (Basal)
Function: Basal subtype marker
[Download JSON]
Component Downloads