CanSys

CanSys is tool that quantifies the contributions of germline and somatic alterations to biological pathway-level disturbances in individual cancer samples. It annotates Variant Call Format (VCF) files with Combined Annotation Dependent Depletion (CADD) scores, calculates gene-level impact scores using CADD and Cancer Dependency Map (DepMap) scores, and maps affected genes onto biological pathways using specified databases (GO and KEGG). We also offer a web-based application cansysplot that allows users to analyze and interactively visualize pathway disturbances in networks. Additionally, the application facilitates the visualization of disturbance networks using data derived from The Cancer Genome Atlas (TCGA).

Requirements

Before running this script, ensure the following software packages and dependencies are installed in your environment.

Software Requirements:

• Python (version 3): Required for running Python scripts involved in the pipeline.
• R (version 4): Required for executing R scripts for statistical analysis.
• tabix: Essential for indexing and rapid retrieval of compressed TAB-delimited genomic data files.

Python Packages:

• pandas: Provides robust data structures for efficient data manipulation and analysis.

R Packages:

• ANNOVAR: Utilized for annotating which genes are associated with variations. Please also download the gene-based annotation file "refGene" (build: hg38).
• dpGMM: Utilized for fitting Gaussian Mixture Model (GMM) to 1D expression data and assigning data points to individual Gaussian components.
• optparse: Facilitates the parsing of command-line options in R scripts.
• fgsea: Fast Gene Set Enrichment Analysis tool.

Setup

CanSys requires no installation. However, users must execute bash download.sh to download the required databases prior to running the CanSys tool. This process may take ～2 hours to complete, and requires ～83GB of available storage. Additionally, users need to specify the ANNOVAR directory path by modifying the annovar_Dir variable in the run_CanSys.sh script.

Usage

./run_CanSys.sh [OPTIONS]

Options:

-Vcf : (Required) Path to the VCF file containing somatic or germline variants for annotation and analysis.

-SampleName : (Required) Name of the sample being analyzed. Used for naming output files.

-Expression : (Optional) Path to the gene expression file. Read count data is recommended. Plasea refer to test/input/expression.txt for an example.

-Output_Dir : (Optional) Specifies the output directory for saving results. If not specified, files will be saved in the current working directory by default.

-Database : (Required) Specifies the database for pathway-level score calculation. Valid options are GO, KEGG, or ALL.

-Cancer : (Required) Specifies the cancer type. Valid options are Ampulla_of_Vater, Biliary_Tract, Bladder_or_Urinary_Tract, Bone, Bowel, Breast, Cervix, CNS_or_Brain, Esophagus_or_Stomach, Eye, Head_and_Neck, Kidney, Liver, Lung, Lymphoid, Myeloid, Ovary_or_Fallopian_Tube, Pancreas, Peripheral_Nervous_System, Pleura, Prostate, Skin, Soft_Tissue, Testis, Thyroid, Uterus and Vulva_or_Vagina.

-Cutoff_CADD : (Optional) Cutoff value for CADD scores. Default: 0.

-nPermSimple : (Optional) Number of permutations used in the permutation test to estimate P-values. Default: 1000.

Examples:

#Run the CanSys tool using the somatic VCF file. Not filter out unexpressed genes.
./run_CanSys.sh -Vcf /path/to/test/input_files/somatic.vcf -SampleName example -Output_Dir /path/to/test/output_files -Database ALL -Cancer Breast

#Run the CanSys tool using the somatic VCF file. Filter out unexpressed genes. Set the number of permutations to 6000.
./run_CanSys.sh -Vcf /path/to/test/input_files/somatic.vcf -SampleName example -Expression /path/to/test/input_files/expression.txt -Output_Dir /path/to/test/output_files -Database ALL -Cancer Breast -nPermSimple 6000

#Run the CanSys tool using the germline VCF file. Filter out unexpressed genes. Filter out variants with CADD scores less then 15.
./run_CanSys.sh -Vcf /path/to/test/input_files/germline.vcf -SampleName example -Expression /path/to/test/input_files/expression.txt -Output_Dir /path/to/test/output_files -Database ALL -Cancer Breast -Cutoff_CADD 15

Outputs

Two output files will be generated for both GO and KEGG analyses: one containing the complete results and the other including only the significant results with Benjamini-Hochberg (BH) adjusted P-values < 0.25. The output files contain detailed information across several columns, each providing specific insights into the biological pathways analyzed. Here is a brief overview of the columns included in the output files:

• goid/entry: This column contains the ID of the biological pathway.
• Name: Lists the names of the biological pathways.
• N: Represents the number of genes within each biological pathway.
• DE: Indicates the number of affected genes present in each biological pathway.
• ES: Refers to the enrichment score.
• NES: Refers to the normalized enrichment score.
• GeneScoreMean: Refers to the average gene-level impact scores in each biological pathway.
• PDS: Refers to the pathway disturbance score (CanSys score).
• pval: Contains the P-values.
• padj: Contains the P-values adjusted using the Benjamini-Hochberg (BH) method.
• Affected Genes: Lists the genes that have a gene-level impact score greater than 0 in each biological pathway.

Citing this work

If you use the CanSys tool or its web-based application cansysplot in your research, please cite: Common and rare germline variants together with somatic mutations alter the integrity of cancer hallmark regulatory networks. (Currently under submission)

Acknowledgments

We would like to express our sincere gratitude to the developers of CADD, DepMap, and all other algorithms and dependencies integrated into our tool.

Contributing

We welcome any bug reports, enhancement requests, and other contributions. To submit a bug report or enhancement request, please use the GitHub issues tracker.

License

This project is licensed under the terms of the MIT license.