Datasets:
Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,66 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
task_categories:
|
| 4 |
+
- tabular-classification
|
| 5 |
+
- regression
|
| 6 |
+
tags:
|
| 7 |
+
- medical
|
| 8 |
+
- oncology
|
| 9 |
+
- synthetic-data
|
| 10 |
+
- genomics
|
| 11 |
+
- ehr
|
| 12 |
+
- nsclc
|
| 13 |
+
- bioinformatics
|
| 14 |
+
pretty_name: ANODE Synthetic Oncology & Genomics
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
# ANODE: Synthetic NSCLC EHR & Genomics Dataset
|
| 18 |
+
|
| 19 |
+
## Title: ANODE-OncoGen High-Fidelity Synthetic Records
|
| 20 |
+
|
| 21 |
+
This dataset consists of a high-density `.jsonl` collection of synthetic patient records for Non-Small Cell Lung Cancer (NSCLC), integrating clinical history with deep genomic markers.
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
|
| 25 |
+
### 📋 Full Feature Specifications
|
| 26 |
+
The dataset includes the following features for every patient record:
|
| 27 |
+
|
| 28 |
+
#### 1. Administrative & Metadata
|
| 29 |
+
* **patient_id**: Unique identifier for the synthetic patient (e.g., `ANODE_551499`).
|
| 30 |
+
* **record_id**: Unique identifier for the specific clinical event.
|
| 31 |
+
* **record_generation_timestamp**: ISO 8601 timestamp of data simulation.
|
| 32 |
+
* **synthetic_source**: Versioning information for the Grok-4 high-fidelity simulator.
|
| 33 |
+
|
| 34 |
+
#### 2. Clinical Demographics
|
| 35 |
+
* **age_at_diagnosis**: Patient age in years.
|
| 36 |
+
* **sex**: Biological sex (male/female).
|
| 37 |
+
* **smoking_pack_years**: Quantitative history of tobacco use.
|
| 38 |
+
* **ecog_performance_status**: Functional status score (0–4) indicating the patient's level of self-care.
|
| 39 |
+
|
| 40 |
+
#### 3. Oncology & Histology
|
| 41 |
+
* **cancer_type**: Specific malignancy (Non-small cell lung cancer).
|
| 42 |
+
* **histology**: Cellular classification (e.g., Adenocarcinoma, Squamous cell carcinoma).
|
| 43 |
+
* **stage_at_diagnosis**: Disease progression stage (e.g., IVA, IVB).
|
| 44 |
+
* **primary_tumor_size_cm**: Diameter of the primary lesion in centimeters.
|
| 45 |
+
* **metastatic_sites_count**: Numerical count of secondary tumor locations.
|
| 46 |
+
|
| 47 |
+
|
| 48 |
+
|
| 49 |
+
#### 4. Genomic & Biomarker Profile
|
| 50 |
+
* **driver_mutation**: Identified genetic drivers (e.g., ALK fusion, EGFR, KRAS, BRAF, or None).
|
| 51 |
+
* **pd_l1_tps_percent**: Tumor Proportion Score for PD-L1 expression (0–100%).
|
| 52 |
+
* **tmb_mutations_per_mb**: Tumor Mutational Burden measured in mutations per megabase.
|
| 53 |
+
* **msi_status**: Microsatellite instability status (e.g., MSS).
|
| 54 |
+
|
| 55 |
+
#### 5. Treatment & Survival Outcomes
|
| 56 |
+
* **first_line_regimen**: Initial therapy used (e.g., carboplatin + paclitaxel + pembrolizumab).
|
| 57 |
+
* **overall_survival_months**: Time elapsed from diagnosis to death or last follow-up.
|
| 58 |
+
* **censored**: Boolean flag (true/false) indicating if the patient was still alive at the end of the simulation period.
|
| 59 |
+
|
| 60 |
+
|
| 61 |
+
|
| 62 |
+
### 📂 Technical Format
|
| 63 |
+
* **File Extension**: `.jsonl` (JSON Lines)
|
| 64 |
+
* **Encoding**: UTF-8
|
| 65 |
+
* **Schema Consistency**: Strict key-value pairing across all entries for programmatic parsing.
|
| 66 |
+
*
|