A Benchmark Study on Error Assessment and Quality Control of CCS Reads Derived from the PacBio RS.

Area of Science:

• Genomics and Bioinformatics
• Molecular Biology

Background:

• Third-generation DNA sequencing technologies, such as PacBio RS, enable the generation of very long DNA reads.
• Assessing sequencing error rates and developing quality control (QC) methods are crucial for new sequencing platforms like PacBio RS.
• Circular Consensus Sequence (CCS) reads are generated by PacBio RS, which are intended to be error-corrected.

Purpose of the Study:

• To evaluate the sequencing error rate of the PacBio RS platform.
• To assess the effectiveness of quality control (QC) methods for PacBio RS sequence data, specifically Circular Consensus Sequence (CCS) reads.
• To determine the impact of QC on downstream bioinformatics applications, such as De Novo assembly.

Main Methods:

• Sequencing of a mixture of 10 known, closely related DNA amplicons using the PacBio RS platform.
• Alignment of generated Circular Consensus Sequence (CCS) reads to known reference sequences.
• Application of a Support Vector Machine (SVM) based multi-parameter QC method to assess and improve read quality.
• Evaluation of different QC approaches using De Novo assembly as a downstream application.

Main Results:

• The median error rate for PacBio RS CCS reads was 2.5% without any read QC.
• Implementation of an SVM-based multi-parameter QC method reduced the median error rate to 1.3%.
• The study demonstrated that appropriate QC is necessary for successful downstream bioinformatics analysis, even with post error-corrected CCS reads.

Conclusions:

• PacBio RS sequencing generates long reads, but quality control is essential for accurate data.
• An SVM-based QC method significantly improves the accuracy of PacBio RS CCS reads.
• Effective QC of CCS reads is critical for reliable downstream bioinformatics analyses, including De Novo assembly.