README.md

---
language:
- en
license:
- mit
task_categories:
- table-to-text
task_ids:
- entity-linking-retrieval
pretty_name: TCGA Cancer Variant and Clinical Data
tags:
- cancer-genomics
- variant-calling
- transcriptomics
- clinical-data
dataset_info:
  features:
  - name: aliquot_id
    dtype: string
  - name: transcript_id
    dtype: string
  - name: mutated_protein
    dtype: string
  - name: wildtype_protein
    dtype: string
  - name: wgs_aliquot_id
    dtype: string
  - name: Cancer Type
    dtype: string
  - name: Cancer Stage
    dtype: string
  - name: Donor Survival Time
    dtype: float
  - name: Donor Vital Status
    dtype: string
  - name: Donor Age at Diagnosis
    dtype: float
  - name: Tumour Grade
    dtype: string
  - name: Donor Sex
    dtype: string
  - name: Histology Abbreviation
    dtype: string
  config_name: default
  splits:
  - name: train
    num_bytes: 193275290
    num_examples: 80840
dataset_files:
- name: protein_sequences_metadata.tsv
  description: >-
    This file contains metadata on protein sequences, including transcript IDs,
    mutated protein sequences, wildtype sequences, and clinical information
    related to cancer studies from TCGA.
  format: tsv
  url: ./protein_sequences_metadata.tsv
size_categories:
- 10K<n<100K
---

# TCGA Cancer Variant and Clinical Data

## Dataset Description

This dataset combines genetic variant information at the protein level with clinical data from The Cancer Genome Atlas (TCGA) project, curated by the International Cancer Genome Consortium (ICGC). It provides a comprehensive view of protein-altering mutations and clinical characteristics across various cancer types.

### Dataset Summary

The dataset includes:

- Protein sequence data for both mutated and wildtype proteins
- Clinical data for each patient, including cancer type, stage, survival time, and demographic information
- Unique identifiers linking genomic data to clinical information

### Supported Tasks and Leaderboards

This dataset can support various tasks in cancer genomics and precision medicine, including:

- Analysis of protein-altering mutations and their impact on cancer progression
- Correlation studies between specific protein mutations and clinical outcomes
- Cancer subtype classification based on genetic and clinical features
- Survival analysis incorporating genetic and clinical data
- Identification of potential biomarkers for cancer prognosis or treatment response

### Languages

The dataset is in English, but primarily consists of protein sequences, numerical data, and standardized clinical terms.

## Dataset Structure

### Data Instances

Each row in the dataset represents a unique combination of a patient sample and a transcript. Here's an example entry:

```sh
{
  'aliquot_id': '8fb9496e-ddb8-11e4-ad8f-5ed8e2d07381',
  'transcript_id': 'ENST00000512632',
  'mutated_protein': 'MACPALGLEALQPLQPEPPPE...', # (truncated for brevity)
  'wildtype_protein': 'MACPALGLEALQPLQPEPPPE...', # (truncated for brevity)
  'wgs_aliquot_id': '80ab6c08-c622-11e3-bf01-24c6515278c0',
  'Cancer Type': 'Liver Cancer - RIKEN, JP',
  'Cancer Stage': '2',
  'Donor Survival Time': 1440.0,
  'Donor Vital Status': 'deceased',
  'Donor Age at Diagnosis': 67.0,
  'Tumour Grade': 'I',
  'Donor Sex': 'male',
  'Histology Abbreviation': 'Liver-HCC'
}
```

### Data Fields

- `aliquot_id`: Unique identifier for the RNA sequencing sample
- `transcript_id`: Ensembl transcript ID
- `mutated_protein`: Amino acid sequence of the mutated protein
- `wildtype_protein`: Amino acid sequence of the wildtype (non-mutated) protein
- `wgs_aliquot_id`: Identifier for the whole genome sequencing data
- `Cancer Type`: Type and origin of the cancer
- `Cancer Stage`: Stage of the cancer at diagnosis
- `Donor Survival Time`: Survival time of the patient in days
- `Donor Vital Status`: Whether the patient is alive or deceased
- `Donor Age at Diagnosis`: Age of the patient at diagnosis
- `Tumour Grade`: Grade of the tumor
- `Donor Sex`: Sex of the patient
- `Histology Abbreviation`: Abbreviation for the cancer histology

### Data Splits

This dataset does not have explicit splits. All data is contained in a single table.

## Dataset Creation

### Source Data

The dataset was derived from the following ICGC/TCGA sources:

1. Normalized transcript expression data:

   ```r
   s3://icgc25k-open/PCAWG/transcriptome/transcript_expression/pcawg.rnaseq.transcript.expr.tpm.tsv.gz
   ```

2. Metadata linked to aliquot_id:

   ```r
   s3://icgc25k-open/PCAWG/transcriptome/metadata/rnaseq.extended.metadata.aliquot_id.V4.tsv.gz
   ```

3. SNV and Indel data:

   ```r
   s3://icgc25k-open/PCAWG/consensus_snv_indel/final_consensus_snv_indel_passonly_icgc.public.tgz
   ```

### Data Processing

The data processing involved several steps:

1. Downloading and extracting the source files
2. Parsing VCF files to extract variant information
3. Translating DNA variants to protein sequences
4. Combining the protein sequence data with clinical data

The script `set_tcga_data.py` was used to perform these processing steps.

## Additional Information

### Dataset Curators

This dataset was curated by [Your Name/Organization] based on data from the ICGC/TCGA Pan-Cancer Analysis of Whole Genomes (PCAWG) project.

### Licensing Information

This dataset is released under the MIT License. Please note that usage of the source TCGA data may be subject to additional terms and conditions.

### Citation Information

If you use this dataset, please cite:

[Your citation information]

And also cite the original PCAWG project:

The ICGC/TCGA Pan-Cancer Analysis of Whole Genomes Consortium. Pan-cancer analysis of whole genomes. Nature 578, 82–93 (2020). https://doi.org/10.1038/s41586-020-1969-6

### Contributions

Thanks to the ICGC/TCGA Pan-Cancer Analysis of Whole Genomes (PCAWG) consortium for making the original data available.