Inferring Variation in Copy Number Using High Throughput Sequencing Data in R.

Brian J Knaus,Niklaus J Grünwald

doi:10.3389/fgene.2018.00123

Abstract

Inference of copy number variation presents a technical challenge because variant callers typically require the copy number of a genome or genomic region to be known a priori. Here we present a method to infer copy number that uses variant call format (VCF) data as input and is implemented in the R package vcfR. This method is based on the relative frequency of each allele (in both genic and non-genic regions) sequenced at heterozygous positions throughout a genome. These heterozygous positions are summarized by using arbitrarily sized windows of heterozygous positions, binning the allele frequencies, and selecting the bin with the greatest abundance of positions. This provides a non-parametric summary of the frequency that alleles were sequenced at. The method is applicable to organisms that have reference genomes that consist of full chromosomes or sub-chromosomal contigs. In contrast to other software designed to detect copy number variation, our method does not rely on an assumption of base ploidy, but instead infers it. We validated these approaches with the model system of Saccharomyces cerevisiae and applied it to the oomycete Phytophthora infestans, both known to vary in copy number. This functionality has been incorporated into the current release of the R package vcfR to provide modular and flexible methods to investigate copy number variation in genomic projects.

Highlights

Investigations into the variation in the number of copies of genes, chromosomes, or genomes are well-established research topics, yet they continue to present technical challenges to molecular genetic analysis
We demonstrate the utility of this method using genomes from the model fungus Saccharomyces cerevisiae and our ongoing work with the oomycete plant pathogen Phytophthora infestans
We developed new functionality added to the current release of the vcfR package that can be used to infer copy number or ploidy in R

Summary

Introduction

Investigations into the variation in the number of copies of genes, chromosomes, or genomes are well-established research topics, yet they continue to present technical challenges to molecular genetic analysis. Whole genome duplication (polyploidy) results in every chromosome being duplicated, a phenomenon observed throughout plants, animals, and fungi (Todd et al, 2017; Van de Peer et al, 2017). This phenomenon is well established, it presents a challenge to high throughput sequencing projects in that most popular genomic variant callers, such as the GATK’s (DePristo et al, 2011) or FreeBayes (Garrison and Marth, 2012), require the a priori specification of how many alleles to call. While the inference of copy number may be an important precursor to point mutation discovery, Inferring Copy Number Variation many authors argue that copy number variation may be more abundant throughout a genome than point mutations (Katju and Bergthorsson, 2013) making it an important facet in the investigation of genomic architectures

Methods

Results

Conclusion