Data Sources

Clinical information

1. coad_clinical.csv: Clinical information for colorectal tumors from the TCGA-COAD dataset.
2. read_clinical.csv: Clinical information for rectal tumors from the TCGA-READ dataset.

These files contain clinical data that we will use to categorize tumors for further analysis.

Description of CSV Files

The two clinical files share similar structures and contain the following important columns:

1. Tumor_Sample_Barcode: A unique identifier for each tumor sample.
2. MSI_TMB: A combined column that contains both the MSI status (Microsatellite Instability) and TMB value (Tumor Mutational Burden). This column is categorizing tumors into the following three groups:
   • MSI-H (microsatellite instability-high)
   • MSS/TMB-H (microsatellite stable with high tumor mutational burden)
   • MSS/TMB-L (microsatellite stable with low tumor mutational burden)

Gene expression

You can download it from google drive

In the gene_expression folder, we have gene expression data for both TCGA-COAD and TCGA-READ cohorts, each located in their respective subfolders:

1. coad/: Contains gene expression data for colorectal tumors.
2. read/: Contains gene expression data for rectal tumors.

Each subfolder includes several files with different types of normalization:

1. counts.csv: Contains the raw gene expression counts for each sample.
2. fpkm.csv: Contains gene expression data normalized using FPKM (Fragments Per Kilobase of transcript per Million mapped reads).
3. fpkm_uq.csv: Contains Upper Quartile (UQ) FPKM normalized data, which adjusts for sequencing depth and variability across samples.
4. tmm.csv: Contains gene expression data normalized using the TMM (Trimmed Mean of M-values) method, often used to correct for library size differences.
5. tpm.csv: Contains gene expression data normalized by TPM (Transcripts Per Million), a method that accounts for gene length and sequencing depth.