Comprehensive evaluation of SNP identification with the Restriction Enzyme-based Reduced Representation Library (RRL)...

Area of Science:

• Genomics
• Molecular Biology

Background:

• Restriction Enzyme-based Reduced Representation Library (RRL) is a feasible method for Single Nucleotide Polymorphism (SNP) identification.
• RRL significantly reduces genome complexity.
• Comprehensive efficacy evaluation of RRL for SNP identification was lacking.

Purpose of the Study:

• To evaluate the efficacy of the Restriction Enzyme-based RRL method for SNP identification.
• To identify factors affecting SNP calling accuracy in RRL.

Main Methods:

• In silico simulation on human reference genome to select Tsp 45I enzyme.
• Sequencing of YH RRL after Tsp 45I treatment.
• SNP identification using SOAPsnp software.
• Comparison with whole genome resequencing data.

Main Results:

• 80.8% of reads mapped to the target region with 20-fold average coverage.
• 96.8% of the target region was covered by at least one read, identifying 257K SNPs.
• False Discovery Rate (FDR) was 13.95% and False Negative Rate (FNR) was 25.90%.
• Concordance rate for homozygote loci was >99.8%, but only 92.56% for heterozygote loci.
• Repeat sequences, base quality, and SNPs in recognition sites impacted accuracy.
• Repeat masking and stringent filters significantly decreased FDR and FNR.

Conclusions:

• Restriction Enzyme-based RRL method is effective for SNP identification.
• Method-derived defects and biases play a crucial role in SNP calling accuracy.
• Careful consideration of repeat masking and stringent filtering criteria is essential for optimizing RRL efficacy.