|Pitfalls and potential of Population genomics|
J. F. Tabima (1), B. J. Knaus (1), N. J. GRUNWALD (2). (1) Oregon State University, Corvallis, OR, U.S.A.; (2) USDA ARS, Corvallis, OR, U.S.A.
<i>Phytophthora</i> pathogens continue to emerge threatening our agricultural and forest systems. New computational and sequencing approaches allow genome-wide characterization of divergence and demographic history in plant pathogen populations. We evaluated a range of approaches including whole genome sequencing, genotyping-by-sequencing (GBS) and Sanger sequencing to infer patterns and processes in the genus <i>Phytophthora</i>. Our experience in calling SNPs revealed that SNP callers (i.e. FreeBayes, SAMtools, GATK and tassel) significantly affected SNP recovery. Furthermore, average read depth needed to be carefully considered during experimental design and implementation. We recommend a strategy, advocated elsewhere, employing a variant discovery phase followed by a variant calling phase, which have goals of high coverage and lower coverage, respectively. While GBS provides huge advantages in terms of cost per SNP and deep SNP coverage across the whole genome, the considerable time investments required to apply sophisticated bioinformatic analysis cannot be ignored. Whole genome sequencing is equally challenging, particularly when applied to hybrids or a population where ploidy varies given difficulties of haplotype phasing. Population genomic approaches combined with high throughput sequencing technologies have the potential to provide novel and unprecedented insights into the emergence of <i>Phytophthora</i> pathogens.