Improved variant calling accuracy by merging replicates in whole-exome sequencing studies.

Area of Science:

• Genomics
• Population Genetics
• Bioinformatics

Background:

• Large-scale population studies often involve whole-exome sequencing (WES).
• Duplicate sequencing of samples can occur for various reasons, presenting challenges for data utilization.
• Efficiently leveraging duplicated WES data is crucial for maximizing study power and accuracy.

Purpose of the Study:

• To evaluate different variant calling strategies for duplicated whole-exome sequencing data.
• To determine the optimal method for utilizing redundant sequencing information in population studies.
• To assess the impact of merging data on variant calling accuracy.

Main Methods:

• Selected 92 samples from a large population study that underwent whole-exome sequencing twice.
• Divided samples into high-depth (H) and low-depth (L) groups, and created a merged group (M).
• Employed the GATK multisample toolkit to compare variant calling accuracy across the three groups, analyzing metrics like Hete/Homo ratio, Ti/Tv ratio, and overlap with the 1000 Genomes Project.

Main Results:

• Hierarchical clustering confirmed high homogeneity between the two sequencing replicates for each subject.
• Variant calling accuracy, assessed by multiple metrics, was consistently higher in the merged group (M) compared to the high-depth (H) and low-depth (L) groups.
• Single nucleotide polymorphisms (SNPs) detected from merged data demonstrated superior data quality.

Conclusions:

• Merging homogeneous duplicated exome sequencing data significantly improves variant calling accuracy.
• This strategy offers a more efficient way to utilize redundant sequencing data in large-scale population studies.
• Combining data from multiple sequencing runs of the same sample enhances the reliability of variant detection.