Home / Pre-computed_databases
Name Modified Size InfoDownloads / Week
Parent folder
Breast_cancer.tgz 2017-07-24 20.3 MB
Lung_adenocarcinoma.tgz 2015-03-24 2.7 MB
Liver_hepatocellular_carcinoma.tgz 2015-03-16 7.0 MB
Ovarian.tgz 2015-03-05 12.4 MB
Prostate_adenocarcinoma.tgz 2015-03-05 9.8 MB
Melanoma.tgz 2015-03-05 20.5 MB
Glioblastoma.tgz 2015-03-05 20.4 MB
Bladder_urothelial_carcinoma.tgz 2015-03-05 15.4 MB
Totals: 8 Items   108.5 MB 0

What is it?

OncoIMPACT is a first-in-class algorithmic framework that nominates patient-specific driver genes by integratively modeling genomic mutations (point, structural and copy-number) and the resulting perturbations in transcriptional programs via defined molecular networks.

OncoIMPACT is configured to run in two modes: (1) a database mode that allows it to determine the parameter settings from the datasets provided (2) a discovery mode where information in the provided database is used to predict driver genes for each sample, in an additional data set.


Installation

Download and uncompress the latest version tar -zxvf oncoIMPACT.tgz chmod +x oncoIMPACT_v0.9.4/*.pl


Usage

Before running the pipeline, please have the following files ready.

Input files and formats

  • CNV_data: a tab-separated file with samples as columns and genes as rows. Each entry of this matrix must hold either a value of -1/0/1. A value of -1 would indicate a DELETION event while a value of 1 will indicate an AMPLIFICATION event, and a value of 0 if the is not affected by copy number gene.

  • SNP_data (point mutations and short indels): a tab-separated file with samples as columns and genes as rows. Each entry of this matrix must hold either the binary values of 0/1. A value of 1 would indicate the presence of a snp in the gene of that particular sample while a value of 0 would indicate otherwise.

  • EXP_data: a tab-separated file with samples as columns and genes as rows. Each entry of this matrix would represent the log2 fold-change of the gene expression of the tumor compared to a normal control.

Config file You will need to create a config file for your project. The config file needs to contain the following parameters (you may refer to the sample config file provided with the test datasets under TEST_DATASET/sampleConfig.cfg).

  • outDir: Full path to destination folder

  • scriptDir: Full path to folder where oncoIMPACT is installed

  • numThreads: Number of threads to use

  • cnv: Full path to cnv data matrix

  • exp: Full path to expression data matrix

  • dataType: Flag for expression data type. Vvalid options: ARRAY (default), RNA_SEQ

  • snp: Full path to snp data matrix

  • dataBase: Full path to the pre-computed database (implies discovery mode only)

  • databaseExport: Full path where the database will be exported (implies database + discovery mode)

  • testMode: Boolean flag to toggle test mode (valid options: 0 / 1)

***at any one time, only one of either the options, <dataBase> or <databaseExport>, may be chosen to run oncoIMPACT.

When you are ready to run the OncoIMPACT pipeline, simply enter the following command oncoIMPACT.pl <path to config file>


Test datasets

We have provided test datasets for running the OncoIMPACT pipeline in 2 different ways.

  • sampleConfig_disc.cfg: sample config for discovery mode

. single_patient.tgz: data for a single GBM patient (this file should be uncompressed)

. Glioblastoma_database.tgz: pre-computed database of the TCGA GBM dataset

  • sampleConfig_db_disc.cfg: sample config for database and discovery mode

. Glioblastoma.tgz: Glioblastoma (GBM) TCGA dataset (this file should be uncompressed)


Change Log

v0.9.4:

  • NEW: Enable error checks

v0.9.3:

  • NEW: Enable the processing of RNA-seq data

  • NEW: Enable the construction of databases

  • NEW: Enable the discovery mode using a pre-computed database

v0.9.2:

  • NEW: Option in configuration file to run OncoIMPACT in test mode which performs the simulation with fewer iterations and fixed seed. In this mode, OncoIMPACT should complete in less than 2 hours using a single thread.

  • NEW: Sanity checks to ensure validity of parameters provided by user

  • FIX: Improved disk space utilization

  • FIX: Improved compatibility with Mac OS.

v0.9.1:

  • oncoIMPACT will now avoid reproducing the input files if COMPLETE_SAMPLES folder exists

  • fix for bugs introduced in last version


Licensing

The MIT License (MIT) Copyright (c) 2014 Genome Institute of Singapore


Contacts

If you want community-driven support, please visit the forum at https://sourceforge.net/p/oncoimpact/discussion/.

If you have a bug to report, you may raise a ticket at https://sourceforge.net/p/oncoimpact/tickets/.

If you have other questions or feedback, you may direct them to Jayce (kohjy@gis.a-star.edu.sg) and Denis (bertrandd@gis.a-star.edu.sg).

Source: README.md, updated 2016-12-09