A catalog of annotated high-confidence SNPs from exome capture and sequencing reveals highly polymorphic genes in Norway spruce (Picea abies)

Abstract

Background: Norway spruce [Picea abies (L.) Karst.] is ecologically and economically one of the most important conifer worldwide. Our main goal was to develop a large catalog of annotated high confidence gene SNPs that should sustain the development of genomic tools for the conservation of natural and domesticated genetic diversity resources, and hasten tree breeding efforts in this species.

Results: Targeted sequencing was achieved by capturing P. abies exome with probes previously designed from the sequenced transcriptome of white spruce (Picea glauca (Moench) Voss). Capture efficiency was high (74.5%) given a high level of exome conservation between the two species. Using stringent criteria, we delimited a set of 61,771 high-confidence SNPs across 13,543 genes. To validate SNPs, a high-throughput genotyping array was developed for a subset of 5571 predicted SNPs representing as many different gene loci, and was used to genotype over 1000 trees. The estimated true positive rate of the resource was 84.2%, which was comparable with the genotyping success rate obtained for P. abies control SNPs recycled from previous genotyping efforts. We also analyzed SNP abundance across various gene functional categories. Several GO terms and gene families involved in stress response were found over-represented in highly polymorphic genes.

Conclusion: The annotated high-confidence SNP catalog developed herein represents a valuable genomic resource, being representative of over 13 K genes distributed across the P. abies genome. This resource should serve a variety of population genomics and breeding applications in Norway spruce.

Keywords: Annotated gene SNPs; Conifer; Exome sequencing; Illumina Infinium iSelect array; Illumina MiSeq; Picea abies; SNP abundance; Single nucleotide polymorphism.

Fig. 1

Pipeline for exome capture and sequencing, sequence assembly and SNP discovery. ¹ [30]. ² [93]. ³ [94]. ⁴454 Life Science, Branford, CT.⁵ See Methods for details

Fig. 2

Distribution of SNP abundance across genes. β is the computed SNP abundance parameter correcting for variable sequencing depth among genes, which was used to compare SNP abundance among them

Fig. 3

Differential representation of gene families among the 10% most SNPed genes versus the other genes. The 16 families found differentially represented after Fisher’s exact test (P < 0.05) are represented; the stars (*) indicate the statistically most significant differences (P < 0.01). Gene family identifiers were retrieved from the ConGenIE database (http://www.congenie.org)