Table 3

Summary of bioinformatics tools for CNV detection using exome sequencing data

Tool

URL

Language

Input

Comments

Ref.


Control-FREECa

http://bioinfo-out.curie.fr/projects/freec/ webcite

C++

SAM/BAM/pileup/Eland, BED, SOAP, arachne, psi (BLAT) and Bowtie formats

Correcting copy number using matched case-control samples or GC contents

[53]

CoNIFERb

http://conifer.sf.net webcite/

Python

BAM

Using singular value decomposition to normalize copy number and avoiding batch bias by integrating multiple samples

[54]

XHMMb

http://atgu.mgh.harvard.edu/xhmm webcite/

C++

BAM

Uses principal component analysis to normalize copy number and HMM to detect CNVs

[55]

ExomeCNVc

http://cran.r-project.org/src/contrib/Archive/ExomeCNV/ webcite

R

BAM/pileup

Using read depth and B-allele frequencies from exome sequencing data to detect CNVs and LOHs

[49]

CONTRAc

http://contra-cnv.sourceforge.net webcite/

Python

SAM/BAM

Comparing base-level log-ratios calculated from read depth between case and control samples

[77]

CONDEX

http://code.google.com/p/condr webcite/

Java

Sorted BED files

Using HMM to identify CNVs

[78]

SeqGene

http://seqgene.sourceforge.net webcite

Python, R

SAM/pileup

Calling variants, including CNVs, from exome sequencing data

[79]

PropSeqc

http://bioinformatics.nki.nl/ocs/ webcite

R, C

N/A

Using the read depth of the case sample as a linear function of that of control sample to detect CNVs

[52]

VarScan2c

http://genome.wustl.edu/software/varscan webcite

Java

BAM/pileup

Using pairwise comparisons of the normalized read depth at each position to estimate CNV

[50]

ExoCNVTestb

http://www1.imperial.ac.uk/medicine/people/l.coin webcite/

Java, R

BAM

Identifying and genotyping common CNVs associated with complex disease

[56]

ExomeDepthb

http://cran.r-project.org/web/packages/ExomeDepth/index.html webcite

R

BAM

Using beta-binomial model to fit read depth of WES data

[30]


aControl-FREEC accepts either matched case-control samples or single sample as input.

bTools use multiple samples as input.

cTools require matched case-control samples as input.

Zhao et al. BMC Bioinformatics 2013 14(Suppl 11):S1   doi:10.1186/1471-2105-14-S11-S1

Open Data