11 May 2009
Recent years have seen rapid advances in the development of technologies that allow massively parallel high-throughput sequencing, referred to as next generation sequencing (NGS) platforms. These platforms have been applied to a number of sequencing studies including examination of RNA-sequencing, whole genome sequencing (see previous news) and genome-wide association studies. The ability of these new sequencing technologies to rapidly generate vast amounts of sequence data at a lower cost makes them suitable for population studies such as the 1000 genomes project (see previous news) and population targeted sequencing studies that look at particular candidate genes and their association with disease. However, sequencing platforms differ considerably in a number of areas including the sequencing chemistry utilised, their technical capabilities such as read length and the methods used for data analysis.
In order to better understand issues that may be faced in data generation and analysis using different sequencing platforms, researchers at the Scripps Research Institute and the Craig J. Ventor Institute evaluated three NGS platforms for targeted sequencing (reported by Genome Web). In their study, the researchers compared the ability of Roche 454, Illumina GA and ABI SOLiD platforms to carry out targeted sequencing of the same 260kb sequence from four individuals [Harismendy et al (2009) Genome Biol. 10(3)R32]. In addition, they compared the data generated by NGS platforms with that generated by traditional Sanger sequencing using an ABI Sanger platform. The authors found that the NGS platforms were able to accurately detect known single nucleotide polymorphism (SNP) variants and had a false negative rate comparable to ABI Sanger. However, the data generated by each of the platforms varied, which was attributed to biases in sample library preparation, variation in sequence coverage depth and systematic errors with each of the platforms. Due to these differences, data from different platforms cannot be used in the same study. The authors recommend optimisation of the uniformity of per-base sequence coverage and reduction of the systematic errors that affect variant calling accuracy in order to effectively balance cost and data quality for population targeted sequencing studies.
Comment: Technological advances have led to the rapid development of a wide array of sequencing platforms. An understanding of the technical aspects of each platform is required in order to identify the best approach for research as well as medical diagnostics applications. Although the platforms used in the above study have been updated since this research was carried out, the paper does allow some insight into the issues that may be encountered when using different sequencing platforms.