Low-density AgriSeq targeted genotyping-by-sequencing markers are efficient for pedigree quality control in Pinus taeda L. breeding

• Published in: Tree genetics & genomes 2023-08, Vol.19 (4), p.34, Article 34
• Main Authors: Lin, Yu-Ming, Shalizi, Mohammad Nasir, Walker, Trevor D., Lauer, Edwin, Carrasco, Claudio D., Gujjula, Krishna Reddy, Suren, Haktan, Isik, Fikret
• Format: Article
• Language: English
• Subjects: Arrays; Biomedical and Life Sciences; Biotechnology; Clustering; Comparative studies; Dairy cattle; Density; Errors; Evergreen trees; Forestry; Genetic analysis; Genetic relationship; Genomes; Genomics; Genotyping; Inbreeding; Life Sciences; Livestock; Nucleotides; Original Article; Pedigree; Pine trees; Pinus taeda; Plant Breeding/Biotechnology; Plant Genetics and Genomics; Principal components analysis; Quality control; Seeds; Single-nucleotide polymorphism; Tree Biology; Trees
• ISSN: 1614-2942; 1614-2950
• DOI: 10.1007/s11295-023-01608-8

Genotyping platforms for breeding programs should be repeatable, reliable, high-throughput, and cost-effective. In this study, we compared the efficiency of an Affymetrix genotyping array (Pita50K) and a targeted genotyping-by-sequencing (tGBS) panel for estimating genomic relationships and detecting potential pedigree errors in a Pinus taeda L. population. The genomic coordinates of 1000 high-quality single nucleotide polymorphic (SNP) markers from the array were used to design primers for amplicon sequencing via AgriSeq™ tGBS technology. The AgriSeq SNP markers were validated through a comparative study. A sample of 192 Pinus taeda individuals with known pedigree were genotyped with the AgriSeq panel as well as the Pita50K array. The two genotyping datasets were compared on the basis of pedigree errors, the contrast between expected and realized genomic relationships, and the genetic clustering of known families. Realized genomic relationships estimated from the AgriSeq panel and the Pita50K genotyping array had a correlation of 0.90. The median (1.077) of inbreeding coefficients from the AgriSeq panel was considerably higher than the median inbreeding coefficients from the pedigree-based (1.000) and SNP array markers-based estimates (1.001). The proportion of significant discrepancy between realized and expected genetic relationships was about 1% in the population. Principal component analysis of realized genetic relationship matrices from two panels showed similar clustering of full-sib families and revealed pedigree errors in the population. The genotyping cost per sample using the AgriSeq platform for P. taeda was one-third of the cost of using the Pita50K SNP array. The platform is a reliable low-density method for genotyping large numbers of samples. Research is underway to test the AgriSeq markers for Mendelian inconsistency and genomic prediction in P. taeda breeding. Other potential research could be marker-assisted backcrossing, F2 enrichment, or prescribed matings between individuals known to carry large-effect resistance genes.