BioXpress Pipeline v-5.0 README

Last Updated August 2021 by Ned Cauley

Description

The BioXpress pipeline takes raw count data from TCGA studies for both Primary Tumor and Normal Tissue and performs differential expression.

The TCGA studies included in the BioXpress pipeline are (by tissue):

• Bladder - TCGA-BLCA (Bladder urothelial carcinoma)

• Breast - TCGA-BRCA (Breast invasive carcinoma)

• Colorectal - TCGA-COAD (Colon adenocarcinoma) - TCGA-READ (Rectum adenocarcinoma)

• Esophageal - TCGA-ESCA (Esophageal carcinoma)

• Head and Neck - TCGA-HNSC (Head and Neck squamous cell carcinoma)

• Kidney - TCGA-KICH (Kidney Chromophobe) - TCGA-KIRP (Kidney renal papillary cell carcinoma) - TCGA-KIRC (Kidney renal clear cell carcinoma)

• Liver - TCGA-LIHC (Liver hepatocellular carcinoma)

• Lung - TCGA-LUAD (Lung adenocarcinoma) - TCGA-LUSC (Lung squamous cell carcinoma)

• Prostate - TCGA-PRAD (Prostate adenocarcinoma)

• Stomach - TCGA-STAD (Stomach adenocarcinoma)

• Thyroid - TCGA-THCA (Thyroid carcinoma)

• Uterine - TCGA-UCEC (Uterine Corpus Endometrial Carcinoma)

Running the Pipeline

To run the BioXpress pipeline, you need to download the scripts from the HIVE Lab github repo: GW HIVE Backend Code Repository If running Bioxpress on a HIVE Lab server (such as glygen-vm-dev), place scripts in your user folder /home/$yourusername/. Data and other output from the scripts is stored in /data/projects/bioxpress/.

Pipeline Overview

Step 1: Downloader

The downloader step will use sample sheets obtained from [GDC Data Portal](https://portal.gdc.cancer.gov/repository) to download raw counts from RNA-Seq for Primary Tumor and Normal Tissue in all available TCGA Studies.

Index for downloader:

• BioXpress Downloader Step
  • General Flow of Scripts
  • Procedure
    • Downloader Step 1 : Get sample list files from the GDC Data Portal
      • Summary
      • Method
    • Downloader Step 2: Run the script get_data_all_samples.sh
      • Summary
      • Method
      • Output
    • Downloader Step 3: get_hits_into_dir.py
      • Summary
      • Method
      • Output
    • Downloader Step 4: Run the script merge_files_tumor_and_normal.sh
      • Summary
      • Method
      • Output

Step 2: Annotation

The annotation step maps transcripts to gene symbols and creates the organized read count and category files used in the DESeq step.

Index for annotation:

• BioXpress Annotation Step
  • General Flow of Scripts
  • Procedure
    • Annotation Step 1 : Run the script merge_per_study.sh
      • Summary
      • Method
      • Output
    • Annotation Step 2 : Run the script merge_per_tissue.py
      • Summary
      • Method
      • Output
    • Annotation Step 3 : Run the script split_per_case.py
      • Summary
      • Method
      • Output

Step 3: DESeq

DESeq is used to calculated differential expression and determine statistical significance.

Index for DESeq:

• BioXpress DESeq step
  • General Flow of Scripts
  • Procedure
    • DESeq step 1: Run the script run_per_study.sh
      • Summary
      • Method
      • Output
    • DESeq Step 2 : Run the script run_per_tissue.sh
      • Summary
      • Method
      • Output
    • DESeq Step 3 : Run the script run_per_case.sh
      • Summary
      • Method
      • Output

Step 4: Publisher

Differential expression results for each tissue are combined into one master dataset.

Index for publisher:

• BioXpress Publisher Step
  • General Flow of Scripts
  • Procedure
    • Publisher Step 1 : Run the script de-publish-per-study.py
      • Summary
      • Method
      • Output
    • Publisher Step 2 : Run the script de-publish-per-tissue.py
      • Summary
      • Method
      • Output

Other documentation

• Major Changes from v-4.0
  • Tumor samples added for each tissue
  • Mapping files updated to reflect most recent mapping of DOIDs to UBERON IDs.
  • Automatic alphabetical re-ordering of count matrices for DESeq2
  • Issue Running DESEq per case
• Post-processing for OncoMX and Glygen
  • Processing for OncoMX
  • Processing for Glygen