electricsheepafrica/colorectal-genomic-ssa

Colorectal Cancer Genomic Profiles (SSA, Synthetic)

Dataset owner: Electric Sheep Africa
Dataset type: Synthetic colorectal cancer genomic profiles (KRAS/BRAF, MSI, Lynch, TMB)
Populations: Colorectal cancer patients from sub-Saharan Africa (modeled cohorts)
Version: 1.0.0
License: CC-BY-NC-4.0

---

1. Dataset Description

This dataset provides synthetic genomic profiles for 30,000 colorectal tumors from sub-Saharan Africa (SSA).

The dataset encodes:

• Per-sample KRAS and BRAF mutation status
• Microsatellite instability (MSI) status and a derived Lynch_syndrome_status label
• Tumor mutational burden (TMB) in mutations per megabase
• Basic clinical context: region, age, sex, tumor_location, and stage

All records are fully synthetic and parameterized using SSA-focused colorectal genomics literature plus global MSI/TMB reviews. No real patient-level data are used.

Important: This dataset contains no real patient data. It is derived entirely from literature-based distributions and internal coherence rules.

---

2. Intended Use

This dataset is intended for:

• Method development for somatic mutation–based risk models and MSI/Lynch detection
• Benchmarking algorithms for TMB-aware modeling and biomarker discovery
• Educational use in colorectal cancer genomics and precision oncology in SSA

Not intended for clinical decision-making or individual patient stratification.

---

3. Cohort Design

3.1 Regions

• West_Africa
• East_Africa
• Southern_Africa
• Central_Africa

3.2 Sample Size & Demographics

• Total samples: 30,000 synthetic colorectal tumors
• Age range: 25–90 years, centered around ~60
• Sex distribution: ~55% Male, 45% Female

3.3 Tumor Location

• Right_colon
• Left_colon
• Rectum

3.4 Stage

• stage ∈ {I, II, III, IV} (enriched for stage II–III)

---

4. Genomic Features

4.1 KRAS and BRAF

Per-sample boolean flags indicate mutation status:

• KRAS_mut
• BRAF_mut

Regional targets are informed by SSA literature:

• West_Africa (e.g., Nigerian cohorts): KRAS ~20–25%; BRAF ~5%
• East_Africa (e.g., Kenyan cohorts): KRAS ~45–50%; BRAF ~5%
• Southern/Central Africa: intermediate values, guided by global CRC ranges

4.2 Microsatellite Instability (MSI) and Lynch Syndrome

• MSI_status ∈ {MSS, MSI-L, MSI-H}
• Lynch_syndrome_status ∈ {Lynch, Non_Lynch}

Approximate targets:

• MSI-H ≈ 15–16% of all CRC
• ~20% of MSI-H tumors are labeled Lynch (≈3% of all CRC), the rest Non_Lynch (sporadic MSI-H)

4.3 Tumor Mutational Burden (TMB)

• tmb_mut_per_mb – Tumor mutational burden (mutations per megabase)

Distributions by MSI class:

• MSS: mean ~4 mut/Mb (low TMB)
• MSI-L: modestly elevated TMB
• MSI-H: mean ~25 mut/Mb (hypermutated), with a wide tail

The validator checks that TMB in MSI-H tumors is several-fold higher than in MSS tumors.

---

5. Main File Schema

5.1 colorectal_genomic_data.csv

• Rows: 30,000 samples
• Columns:

Clinical & context:

• sample_id – Synthetic ID (CRC_SAMPLE_00000 ...)
• region – {West_Africa, East_Africa, Southern_Africa, Central_Africa}
• age – Age in years
• sex – {Male, Female}
• tumor_location – {Right_colon, Left_colon, Rectum}
• stage – {I, II, III, IV}

Genomic features:

• MSI_status – {MSS, MSI-L, MSI-H}
• Lynch_syndrome_status – {Lynch, Non_Lynch}
• tmb_mut_per_mb – Tumor mutational burden (mut/Mb)
• KRAS_mut – Boolean, KRAS mutated
• BRAF_mut – Boolean, BRAF (primarily V600E) mutated

---

6. Data Access & Files

Main Dataset (root)

• colorectal_genomic_data.csv – 30,000 × 11 variables (main table, CSV)
• colorectal_genomic_data.parquet – Same table in Parquet format for efficient loading

No auxiliary long-format variant or CNV tables are provided for this dataset; it focuses on per-sample biomarker status and TMB.

---

7. Generation & Validation (Summary)

The dataset was generated using the GENOMICS Synthetic Data Playbook:

1. Domain specification – Identify key colorectal biomarkers (KRAS, BRAF, MSI, Lynch, TMB) relevant for SSA cohorts.
2. Literature-informed parameters – Extract KRAS/BRAF frequencies from Nigerian and Kenyan cohorts, MSI-H and Lynch fractions from South African and global CRC studies, and TMB distributions from MSI/TMB reviews (see local docs/LITERATURE_INVENTORY.csv).
3. Configuration – Encode parameters in colorectal_genomic/configs/colorectal_genomic_config.yaml.
4. Generation – Use scripts/generate_colorectal_genomic.py to simulate 30,000 tumors.
5. Validation – Use scripts/validate_colorectal_genomic.py to check distributions, MSI/Lynch prevalence, TMB patterns, and missingness.

7.1 Validation Status

• Validation run on: colorectal_genomic/output/colorectal_genomic_data.csv
• Checks include:
  • Region distribution vs config
  • KRAS/BRAF prevalences (overall and by region) vs config
  • MSI and Lynch prevalence vs config
  • TMB shift in MSI-H vs MSS
  • Missingness in key variables

The current validation report indicates Overall Status: PASS (all checks PASS).

---

8. Example Usage

8.1 Load with pandas

import pandas as pd

df = pd.read_csv("colorectal_genomic_data.csv")
print(df.shape)
print(df.head())

8.2 Load with datasets (Hugging Face)

from datasets import load_dataset

crc = load_dataset("electricsheepafrica/colorectal-genomic-ssa")
df = crc["train"].to_pandas()

print(f"Samples: {len(df)}")
print(f"Columns: {len(df.columns)}")

8.3 TMB vs MSI status

summary = df.groupby("MSI_status")["tmb_mut_per_mb"].describe()
print(summary)

8.4 KRAS/BRAF by region

region_mut = df.groupby("region")["KRAS_mut", "BRAF_mut"].mean()
print(region_mut)

---

9. Ethical & Appropriate Use

9.1 Appropriate Uses ✅

• Method development and benchmarking in CRC biomarker modeling
• Educational use in colorectal cancer genomics and TMB/MSI modeling
• Health equity and disparity analysis using synthetic data

9.2 Inappropriate Uses ❌

• Clinical decision-making for individual patients
• Real-world prognosis or treatment selection
• Insurance, employment, or financial decisions

This dataset is synthetic and must not be used as a substitute for real clinical or genomic data in patient care.

---

10. License

• License: CC-BY-NC-4.0
• Commercial use: Not permitted without explicit permission.

---

11. Citation

Electric Sheep Africa (2025).
Colorectal Cancer Genomic Profiles (SSA, Synthetic Dataset).
Generated using literature-informed synthetic data methods.
Hugging Face Datasets. Version 1.0.0.

---

12. Contact

• Organization: Electric Sheep Africa
• Hugging Face: https://huggingface.co/electricsheepafrica

Feedback and collaboration inquiries are welcome.