TitanCNA a R/Bioconductor package for analyzing subclonal copy number alterations (CNA) and loss of heterozygosity (LOH) in whole genome and exome sequencing of tumours.
Ha, G., et al. (2014). TITAN: Inference of copy number architectures in clonal cell populations from tumour whole genome sequence data. Genome Research, 24: 1881-1893. (PMID: 25060187)
Gavin Ha
Fred Hutchinson Cancer Research Center
contact: [email protected] or [email protected]
Date: May 30, 2019
Website: GavinHaLab.org
Snakemake Workflow: https://github.com/gavinha/TitanCNA/tree/master/scripts/snakemake
10X Snakemake Workflow: https://github.com/gavinha/TitanCNA_10X_snakemake
Google Groups: https://groups.google.com/forum/#!forum/titancna
Publication in Genome Research: http://genome.cshlp.org/content/24/11/1881
(See NEWS for previous version notes)
Addressed the issue of RangedData
being deprecated by converting code to use GRanges
from the GenomicRanges
package.
New or modified functions:
- wigToGRanges
: to load WIG files and store in GRanges
object.
- correctIntegerCN()
: performs allelic copy number (major/minor CN) adjustment.
Improved parameter inference by handling errors and allowing EM to continue until convergence. This fixes runs that previously would fail because samples had very low tumor content.
Snakemake workflow for 10X Genomics whole genome sequencing data is now included in another Git repo. https://github.com/gavinha/TitanCNA_10X_snakemake
-
New functions:
correctIntegerCN()
: recomputes high-level copy number that is capped by the maximum CN state. Performs two tasks - (1) correct log ratio based on purity and ploidy, and then convert to decimal CN value; (2) Correct bins and segments in which the original predicted integer copy number was assigned the maximum CN state; bins and segments for all of chromosome X are also corrected, if provided in the input.
-
Modified functions:
plotSegmentMedians()
andplotCNlogRByChr()
: includes argument to show color-coding for corrected copy number; defaults to TRUE for this argument.
-
Removed functions/manual/dependencies:
extractAlleleReadCounts()
Rsamtools
dependency
- 10X Genomics analysis
- Please see scripts for instructions on running the 10X Genomics analysis.
-
New script to help select optimal solutions.
Please see scripts/R_scripts -
Added snakemake pipeline for entire TITAN workflow
Please see scripts/snakemake. -
New function:
plotSegmentMedians()
loadHaplotypeAlleleCounts()
: loads input allele counts with phasing informationplotHaplotypeFraction()
: results from 10X Genomics WGS data with phasing of haplotype blocks
-
Modified features (no changes for user-accessible functions):
- updateParameters: coordinate descent estimate of ploidy update uses previously estimated normal parameter from the same corodinate descent iteration ; leads to faster convergence
From within R-3.3.2 or higher,
install.packages("devtools")
library(devtools)
install_github("gavinha/TitanCNA")
From within R-3.3.2 or higher,
if (!requireNamespace("BiocManager", quietly=TRUE))
install.packages("BiocManager")
BiocManager::install("TitanCNA")
-
Install the HMMcopy suite Please follow instructions on the HMMcopy GitHub https://github.com/shahcompbio/hmmcopy_utils.
-
Install ichorCNA Please follow instructions on the ichorCNA GitHub Wiki https://github.com/broadinstitute/ichorCNA.
R scripts are provided to run the R component of the TITAN analysis using the TitanCNA R/Bioconductor package.
Please go to the scripts directory and look at the README there for more details.
A snakemake is also provided in this repo.
This workflow will run the TITAN a set of tumour-normal pairs, starting from the BAM files and generating TitanCNA outputs. It will also perform model selection at the end of the workflow to choose the optimal ploidy and clonal cluster solutions.
The PDF of the vignette can be accessed from R
library(TitanCNA)
browseVignettes(package = "TitanCNA")
The path of the file can also be located using
pathToInstall <- system.file(package = "TitanCNA")
pathToPdf <- paste0(pathToInstall, "/int/doc/TitanCNA.pdf)
The example provided will reproduce Figure 1 in the manuscript. However, it will be slightly different because the example is only based on the analysis of chr2, not genome-wide.
TitanCNA was developed by Gavin Ha while in the laboratories of Sohrab Shah ([email protected]) and Sam Aparicio ([email protected]) at the Dept of Molecular Oncology, BC Cancer Agency, Vancouver, Canada.
Yikan Wang and Daniel Lai have contributed code and discussions to this project.
The KRONOS TITAN workflow was developed by Diljot Grewal ([email protected]) and Jafar Taghiyar ([email protected]).
HMMcopy was co-developed by Daniel Lai and Gavin Ha.
TitanCNA was inspired by existing methods including OncoSNP and PyClone
License: GPLv3
TitanCNA R code is open source and R/Bioconductor package is under GPLv3. This applies to the v1.9.0 and all subsequent versions within and obtained from Bioconductor.
Users who are using TitanCNA earlier than v1.9.0 not for the purpose of academic research should contact [email protected], [email protected], and [email protected] to inquire about previous licensing.