GUST - Genes Under Selection in Tumors

GUST predicts oncogenes (OGs), tumor suppressor gens (TSGs) and passenger genes (PGs) using somatic SNVs detected in a collection of tumors. It can classify single genes from targetted sequencing or multiple genes from whole-exome sequencing.

Install.

# Be patient. It may take some time.
 devtools::install_github('liliulab/gust')

Usage

library(gust)

# to make predictions. 
# A sample input can be downloaded from the "examples" folder. 
 gust(input.file.name='./examples/TCGA.ACC.mutect.somatic.maf.gz', output.folder='./examples/', output.prefix='TCGA.ACC');
# Compare outputs from the above command with files in the "examples" folder.
 
# to plot distribution
 m=plot.this.one('TP53', './examples/', 'TCGA.ACC') 

The input file needs to be in the VCF format. Required fields include Tumor_Sample_Barcode, Chromosome, Start_Position, dbSNP_RS, Reference_Allele, Tumor_Seq_Allele2, FILTER, One_Consequence, Hugo_Symbol, Gene, Feature, ENSP, HGVSc, HGVSp_Short, Amino_acids, Codons, ENSP, RefSeq, and Entrez_Gene_Id. We recommend using SnpEff to annotate somatic variants, which will generate a VCF file containing the required fields. Additional usages can be found via using "?gust" after installing the R gust package.

Precomputed results

GUST classifications of genes in 33 cancer types using the TCGA whole-exome sequencing data are available at https://www.liliulab.com/gust.

Reference

Please cite our publication in Bioinformatics.

Chandrashekar P, Ahmadinejad N, Sekulic A, Wang J, Kumar S, Maley C, Liu L. (2019) Somatic selection distinguishes oncogenes and tumor suppressor genes. Bioinformatics (in press) (advanced online publication https://academic.oup.com/bioinformatics/advance-article/doi/10.1093/bioinformatics/btz851/5625621)

Contributors

Algorithm of GUST was developed by Li Liu. Database was developed by Pramod Chandrashekar. Please contact liliu at asu.edu for any questions or suggestions.