VarDict

VarDict is an ultra sensitive variant caller for both single and paired sample variant calling from BAM files. VarDict implements several novel features such as amplicon bias aware variant calling from targeted sequencing experiments, rescue of long indels by realigning bwa soft clipped reads and better scalability than many Java based variant callers.

Due to the philosophy of VarDict in calling "everything", several downstream strategies have been developed to filter variants to for example the most likely cancer driving events. These strategies are based on evidence in different databases and/or quality metrics. http://bcb.io/2016/04/04/vardict-filtering/ provides an overview of how to develop further filters for VarDict. The script at https://github.com/AstraZeneca-NGS/VarDict/blob/master/vcf2txt.pl can be used to put the variants into a context by including information from dbSNP, Cosmic and ClinVar. We are open to suggestions from the community on how to best narrow down to the variants of most interest.

A Java based drop-in replacement for vardict.pl is being developed at https://github.com/AstraZeneca-NGS/VarDictJava. The Java implementation is approximately 10 times faster than the original Perl implementation and does not depend on samtools

To enable amplicon aware variant calling (single sample mode only; not supported in paired variant calling), please make sure the bed file has 8 columns with the 7th and 8th columns containing the insert interval (therefore subset of the 2nd and 3rd column interval). The bed files typically look similar to the below two overlapping intervals:

chr1 115247094 115247253 NRAS 0 . 115247117 115247232

chr1 115247202 115247341 NRAS 0 . 115247224 115247323

For more information on amplicon aware calling please see https://github.com/AstraZeneca-NGS/VarDict/wiki/Amplicon-Mode-in-VarDict

VarDict is fully integrated in e.g. bcbio-nextgen, see https://github.com/chapmanb/bcbio-nextgen

Please cite VarDict:

Lai Z, Markovets A, Ahdesmaki M, Chapman B, Hofmann O, McEwen R, Johnson J, Dougherty B, Barrett JC, and Dry JR. VarDict: a novel and versatile variant caller for next-generation sequencing in cancer research. Nucleic Acids Res. 2016, pii: gkw227.

The link to is article can be accessed through: http://nar.oxfordjournals.org/cgi/content/full/gkw227?ijkey=Tk8eKQcYwNlQRNU&keytype=ref

Coded by Zhongwu Lai 2014.

Requirements

• Perl (uses /usr/bin/env perl)
• R (uses /usr/bin/env R)
• samtools (must be in path, not required if using the Java implementation in place of vardict.pl)

Quick start

Make sure the VarDict folder (scripts vardict.pl, vardict, testsomatic.R, teststrandbias.R, var2vcf_valid.pl and var2vcf_somatic.pl) is in path before running the following commands.

• Running in single sample mode:

  AF_THR="0.01" # minimum allele frequency
  vardict -G /path/to/hg19.fa -f $AF_THR -N sample_name -b /path/to/my.bam -c 1 -S 2 -E 3 -g 4 /path/to/my.bed | teststrandbias.R | var2vcf_valid.pl -N sample_name -E -f $AF_THR
  

• Paired variant calling:

  AF_THR="0.01" # minimum allele frequency
  vardict -G /path/to/hg19.fa -f $AF_THR -N tumor_sample_name -b "/path/to/tumor.bam|/path/to/normal.bam" -c 1 -S 2 -E 3 -g 4 /path/to/my.bed | testsomatic.R | var2vcf_somatic.pl -N "tumor_sample_name|normal_sample_name" -f $AF_THR
  

Contributors

• Zhongwu Lai, AstraZeneca

• Miika Ahdesmaki, AstraZeneca

• Brad Chapman, Harvard School of Public Health

License

The code is freely available under the MIT license.