Variant calling and genotyping accuracy of ddRAD-seq: Comparison with 20X WGS in layers

Area of Science:

• Animal Genomics
• Molecular Biology
• Quantitative Genetics

Background:

• Whole Genome Sequencing (WGS) is currently too expensive for routine laying hen genotyping.
• SNP chips are commonly used but alternative, cost-effective sequencing methods are needed.
• Reduced representation sequencing approaches balance quality and cost by targeting specific genomic regions.

Purpose of the Study:

• To evaluate double digested Restriction site Associated DNA sequencing (ddRAD-seq) for SNP identification and genotyping in laying hens.
• To compare ddRAD-seq performance against Whole Genome Sequencing (WGS).
• To assess the impact of data filtration thresholds on genotyping reliability and SNP quantity.

Main Methods:

• ddRAD-seq was performed on laying hens and compared to WGS data.
• Variant calling sensitivity, precision, and genotyping reliability were determined.
• SNP Call Rate (CRSNP) and mean depth of sequencing per SNP (DPSNP) were analyzed.

Main Results:

• Raw ddRAD-seq identified 349,497 SNPs with 80% mean genotyping reliability.
• Sensitivity and precision for SNPs within expected fragments were 32.4% and 96.4%, respectively.
• SNPs outside expected fragments showed significantly lower CRSNP and DPSNP, impacting overall reliability.

Conclusions:

• ddRAD-seq is a suitable method for SNP variant calling and genotyping in laying hens.
• Establishing filtration thresholds for CRSNP and DPSNP is crucial for optimizing genotyping reliability and SNP yield.
• Severe quality control on ddRAD-seq data can yield a minimum of 40% of SNPs with 98% reliability.