Amplicon-based semiconductor sequencing of human exomes: performance evaluation and optimization strategies

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• The Ion Proton platform offers cost-effective whole exome sequencing (WES) with rapid turnaround.
• AmpliSeq Exome kits and HiQ sequencing chemistry are key components for WES on this platform.
• Accurate variant identification is crucial for clinical applications of WES.

Purpose of the Study:

• To assess the performance of Ion Proton WES for variant identification using gold-standard data.
• To characterize erroneous variant calls and develop a filtering strategy to minimize false positives.
• To evaluate the impact of sequencing depth and capture efficiency on variant detection.

Main Methods:

• Whole exome sequencing was performed on the Ion Proton platform.
• The AmpliSeq Exome kit was used for library preparation.
• Performance was evaluated using variants from the Genome in a Bottle (GIAB) consortium.
• Variant identification accuracy was assessed at different sequencing coverages.
• A filtering strategy was developed and applied to reduce false-positive variant calls.

Main Results:

• The AmpliSeq Exome kit captures over 94% of human coding bases, with some genes showing incomplete representation.
• Optimal variant detection performance was achieved at mean coverage >120×.
• HiQ chemistry demonstrated sensitivities of ~71% for indels and ~97.5% for SNPs.
• The proposed filtering strategy reduced false positives by up to 40.4% for indels and 68.2% for SNPs.
• Higher sequencing coverage (>120×) improved variant detection accuracy.

Conclusions:

• Amplicon-based WES on the Ion Proton platform with HiQ chemistry is a competitive approach for variant identification.
• While false positives remain a challenge for Ion Torrent technology, the developed filtering strategy effectively reduces erroneous calls.
• The platform shows promise for cost-effective and accurate genetic variant analysis, particularly with optimized sequencing depth and filtering.