A 3K Axiom^® SNP array from a transcriptome-wide SNP resource sheds new light on the genetic diversity and structure of the iconic subtropical conifer tree Araucaria angustifolia (Bert.) Kuntze

Abstract

27High-throughput SNP genotyping has become a precondition to move to higher precision 28 and wider genome coverage genetic analysis of natural and breeding populations of non-29 model species. We developed a 44,318 annotated SNP catalog for Araucaria angustifolia, a 30 grandiose subtropical conifer tree, one of the only two native Brazilian gymnosperms, 31 critically endangered due to its valuable wood and seeds. Following transcriptome assembly 32 and annotation, SNPs were discovered from RNA-seq and pooled RAD-seq data. From the 33 SNP catalog, an Axiom® SNP array with 3,038 validated SNPs was developed and used to 34 provide a comprehensive look at the genetic diversity and structure of 15 populations across 35 the natural range of the species. RNA-seq was a far superior source of SNPs when 36 compared to RAD-seq in terms of conversion rate to polymorphic markers on the array, 37 likely due to the more efficient complexity reduction of the huge conifer genome. By 38 matching microsatellite and SNP data on the same set of A. angustifolia individuals, we 39 show that SNPs reflect more precisely the actual genome-wide patterns of genetic diversity 40 and structure, challenging previous microsatellite-based assessments. Moreover, SNPs 41 corroborated the known major north-south genetic cline, but allowed a more accurate 42 attribution to regional versus among-population differentiation, indicating the potential to 43 select ancestry-informative markers. The availability of a public, user-friendly 3K SNP array 44 for A. angustifolia and a catalog of 44,318 SNPs predicted to provide ~29,000 informative 45 SNPs across ~20,000 loci across the genome, will allow tackling still unsettled questions on 46 its evolutionary history, toward a more comprehensive picture of the origin, past dynamics 47 3 48 SNP array described, unlocks the potential to adopt genomic prediction methods to 49 accelerate the still very timid efforts of systematic tree breeding of A. angustifolia. 50 51